Agentic AI for IT Operations: Windows 365, Intune & Microsoft's AI Vision with Christiaan Brinkhoff [Nerdio]
Explore agentic AI for IT operations with Christiaan Brinkhoff: Windows 365, Intune, AI agents, governance, security, and Microsoft's agentic OS vision.
Agentic AI for IT Operations: Windows 365, Intune & Microsoft's AI Vision with Christiaan Brinkhoff [Nerdio]
Quick answer: Agentic AI for IT operations combines goal-driven AI agents with governed workflows, observability, identity, and human oversight. In this episode, Christiaan Brinkhoff explains how Windows 365, Microsoft Intune, Agent 365, Microsoft Foundry, and the emerging agentic operating system can reshape endpoint management, security, and cloud operations.
What You Will Learn
- How agentic AI differs from generative AI, chatbots, and traditional automation.
- Where Windows 365 and Intune can support autonomous IT operations and Cloud PC management.
- How Agent 365, Microsoft Foundry, Microsoft Entra, Microsoft Purview, and Defender contribute to governance and security.
- How to introduce autonomy gradually through guardrails, observability, measurable outcomes, and human oversight.
Agentic AI for IT Operations: A Practical Definition
Agentic AI for IT operations describes AI systems that can observe context, plan work, use tools, execute actions, and adapt to results while pursuing a defined objective. Unlike a conventional script, an agent can evaluate changing conditions; unlike a chatbot, it can carry out approved actions. In enterprise environments, autonomy must remain bounded by identity, permissions, policy, logging, and escalation paths.
Artificial intelligence is rapidly transforming IT operations, with "agentic systems" leading the next wave of autonomous, intelligent innovation. Driven by this evolution, the global AIOps market is projected to expand at a compound annual growth rate (CAGR) of 15.2% from 2025 to 2030. Microsoft's AI Vision aligns perfectly with this paradigm shift, seamlessly connecting Windows 365 and Microsoft Intune to redefine IT management, elevate enterprise security, and deliver superior user experiences. Organizations adopting agentic AI can drastically reduce operational overhead and cut manual administrative workloads by up to 80%.
Agentic AI enables powerful self-service capabilities that dramatically lower support expenses. By reducing the cost of individual support interactions to under $2, cloud-driven autonomous workflows offer significant financial advantages over traditional human-assisted helpdesk models.

How will agentic AI revolutionize modern IT orchestration, and how does Microsoft's AI Vision shape the future of cloud computing?
Key Takeaways
- Agentic AI systems fix hard problems by themselves. They cut IT costs by 80 percent.
- Microsoft has tools like Agent 365 and Foundry. These help companies make safe AI agents easily.
- Windows 365 and Intune use automatic security agents. They keep devices safe. They stop cyber threats quickly.
- Microsoft is making Windows an Agentic OS. It will do daily work tasks automatically. This will make people more productive.
Agentic Systems: The IT Paradigm Shift
Defining Autonomous AI Agents
Agentic systems are a new kind of artificial intelligence. These AI agents work on their own. They solve hard problems. They learn from their surroundings. This helps them reach their goals. They do not just follow orders. They understand situations. They change what they do. This makes them different. Older automation did not do this.
| Feature Dimension | Traditional Automation / AI | Agentic Systems |
|---|---|---|
| Behavioral Intelligence | Acts only when told. | Watches things all the time. Knows what is needed next. |
| Workflow & Planning | Does simple things. Fails at hard tasks. | Handles many steps. Remembers what happened. Changes plans as needed. |
| Learning & Adaptability | Uses old information. Needs full restart to update. | Learns all the time. Gets better without big changes. |
| Domain Flexibility | Made for one small job. Cannot do other things well. | Is smart in general. Works well in new situations. |
Old systems use fixed rules. People always tell them what to do. But agentic systems think for themselves. They break big goals into small jobs. They use tools to finish these jobs. These AI agents work independently. They change their plans. This is based on new information. It also depends on changes around them.
Transforming IT Operations with Agentic AI
Agentic AI changes IT operations. It is more than simple automation. It makes your work smart. It makes it flexible. This change affects many IT areas.
| IT Operational Area | Agentic AI Impact / Capability |
|---|---|
| Self-Service | Uses talking AI. |
| Service Desk | Uses smart automation. |
| Insights & Analytics | Gets better with smart predictions. |
| Infrastructure & Operations | Fixed automatically. |
| Asset Management | Managed automatically. |
You will see better self-service. Your service desk works better. This is with smart automation. Smart predictions improve your data. Automatic fixes manage your systems. Automatic handling makes asset management better. These AI agents and apps make your IT more active.
Why Agentic Systems are Crucial for Enterprises
Agentic systems are key for businesses today. They offer big business benefits. You get more efficient work. Your employees do more. These systems help businesses grow easily. They make decisions faster. Your operations become very quick. They help find new ways to help customers. You also save money. This is by changing how work is done. Agentic AI improves important tasks. These include risk management. It also helps with supply chain strength.
Microsoft's Agentic AI Frameworks
You need a strong base. This base helps manage new AI agents. Microsoft gives you full frameworks. These frameworks make sure your AI is safe. They make sure it follows rules. They also make sure it is well-managed. They give you one place to control all your AI agents.
Agent 365: The Control Plane for AI Agents
Agent 365 is your main spot. It manages AI agents. You can see everything. You can control everything. This system helps you watch all your AI agents. They are across Microsoft. You can see how agents relate. You can see how well they work. It checks how fast they work. It checks business returns. Agent 365 also makes sure agents follow rules. It makes sure they are safe from the start. It stops agents from using too much. This is for certain users or data. It makes sure they only use what they need. Agent 365 tracks and logs things. This makes it clear. It keeps agent identities safe. It uses Microsoft Entra. It finds weak spots. It stops attacks. It uses Microsoft Defender. This system uses Microsoft Purview. This is for data safety. It stops AI agents from sharing too much data.
Microsoft Foundry for Agent Development
Microsoft Foundry helps you build AI agents. It helps you use them. It has over 11,000 models. You can pick from them. You can make smart agents. They know what is happening. Foundry gives you server space. This is for hosting agents. It connects agents to your company's knowledge. This system lets agents act on their own. This is across other systems. Foundry also gives you one way to manage. It stops security threats. It lets you see things right away. This helps you manage your AI agents well.
Centralized Governance with Entra ID and M365 Admin Center
Entra ID and M365 Admin Center control your AI agents. They are one main source of truth. This has much information for agents. It is across Microsoft. It is also outside Microsoft. You can group agents. This is by how much you trust them. Or by what they do. This applies rules to many agents. Each AI agent gets its own identity. This lets non-human things log in. It also makes sure security rules are followed. You can see everything in one place. Every AI agent made is listed. This helps you manage your security.
Compliance and Data Protection via Purview
Microsoft Purview keeps AI agent data safe. It follows rules. It tracks who looks at data. It flags private things. This is in agent results. Purview warns you about strange data flows. It finds when rules are broken. You get full records of data used. It sets strict limits on permissions. Purview also looks for attack paths. It shows where data might be exposed. This helps stop data loss.
Unified Security with Defender XDR
Microsoft Defender XDR keeps your AI safe. It finds weak spots. It stops bad attacks. It fixes security problems. Microsoft Defender helps keep your AI agents safe. It works with Microsoft Sentinel. This gives you full threat protection. You get advanced protection for AI. This makes your security stronger.
You get a full Windows experience. It is personal. It comes from the cloud. This is with Windows 365. This platform is fully managed. It removes hard VDI problems. You can set up Cloud PCs fast. You can secure them. You can use them quickly. Users get easy access. This is across many devices. Windows 365 works with Microsoft Intune. This integration lets you manage both. You manage physical and Cloud PCs together. You use the same security rules. You use the same compliance checks.
Intelligent Endpoint Management with Intune
Intune manages all your endpoints. It manages physical devices. It also manages Cloud PCs. You use the same security rules. You use the same work steps. Intune uses AI. This makes management smarter. It is more than simple automation. AI agents in Intune work alone. They learn. They change to fit your system.
Think about these smart AI agents in Intune:
| Intelligent AI Agent | Functionality / Capability |
|---|---|
| Change Review Agent | Checks proposed changes. It does this automatically. |
| Policy Configuration Agent | Makes setting up rules faster. |
| Device Offboarding Agent | Handles removing devices. It does this safely. |
These agents help manage devices better. They keep your security posture strong.
Enhancing Cloud PCs with Windows 365 AI
Windows 365 gives personal Windows experiences. AI makes these Cloud PCs much better. Features like Windows 365 Boot exist. Users log right into their cloud desktop. This happens when they start their device. Windows 365 Switch allows easy moving. You move between local and cloud. These features make users happier. AI makes Cloud PCs work better. It uses resources well. You always get the best experience. This makes your Cloud PCs faster. They are also more reliable.
Windows 365 for Agents: Automated Workflows
Windows 365 for Agents gives secure Cloud PCs. They are for AI agents. It has built-in viewing. This platform helps manage your agents. It manages their whole life.
- End-to-End Lifecycle Management: You control sessions. You control networking. You control capacity. You control data location limits.
- Observability & Real-Time Monitoring: You see live pictures. You can take control directly. Audit screenshots have timestamps. They check security.
- Pay-As-You-Go Pricing: This lowers running costs. You pay only for active use. This is good for changing agent work.
- Cross-Platform OS Support: It works with Windows. It works with Linux. It works with browsers. This runs many types of automation.
Windows 365 for Agents is a strong place. It is for your AI work.
| Feature Category | Implementation Details | Operational Benefit |
|---|---|---|
| Execution Environment | Managed Windows Cloud PCs. They run on Azure VM. | Allows agents to work alone. They can browse the web. They can process data. This is inside safe company limits. |
| Management & Identity | Works with Microsoft Intune. This is for device control. Works with Microsoft Entra ID. This is for identity. | Makes sure rules are followed. Gives strong access security. No manual setup is needed. |
| Resource Allocation | Shared Cloud PC groups. They are by team or work. They have ready or planned capacity. | No strict one-to-one rules. Allows changes based on tasks. |
This setup makes your AI agents work well. They work safely.
Proactive Security Agents for Threat Response
AI agents are key for security. They act early against threats. This is in Windows 365 and Intune. These agents watch your system always. They look for strange things. When they find a threat, they act fast. Microsoft Defender XDR works with these agents. This gives strong threat protection. It makes your Windows security better. These agents can cut off bad devices. They can block bad traffic. They can start automated incident response. This means threats are found faster. They are stopped faster.
Streamlining Identity and Access with AI
AI makes identity and access easier. It works across Windows 365 and Intune. AI agents can set up users automatically. They can remove users automatically. They enforce access rules. This is based on the situation. For example, an AI agent can see odd logins. It can then ask for more proof. This makes your compliance stronger. It also makes overall security better. You manage user identities better. This ensures only allowed users. They access your Cloud PCs and data.
Microsoft's AI Vision: The Agentic OS
Microsoft's AI vision is changing Windows. It will become an "Agentic OS." This means AI will do hard tasks. It goes beyond simple chatbots. This is a big step forward. You will get a smarter system. It will be more active.
Windows Evolution as an AI-Native Platform
Microsoft sees Windows as an AI-native platform. This means AI is built in. Key changes make this happen. The Model Context Protocol (MCP) is part of the system. This open system lets AI agents connect. They link with apps. They link with tools. They link with system tasks. This automates daily work for you. The Windows On-Device Registry (ODR) also helps. This safe system stores agent links. It gives you control. Native Agent Connectors are built-in. They work with File Explorer. They work with System Settings. They let agents change settings. They manage files. They help you. You always stay in control. Cloud management with Intune builds a base. This supports new AI. It gets your systems ready.
Copilot's Role in the Agentic Ecosystem
Copilot is key in this AI system. It is an AI helper. It makes you more productive. It helps you be creative. It gives live help. It gives ideas. It gives useful tips. Copilot is the main chat tool. You use it to talk to many AI agents. These include agents for Sales. Also for Service. And for Finance. Copilot makes digital work better. This is in Word. Also in Outlook. And in Teams. And in Copilot Chat. It uses company info. This is from Microsoft Graph. It also provides AI tools. This includes basic models. It includes systems to run agents. These agents work safely. They follow company rules.
Cloud-Powered Flexibility and AI Optimization
Cloud power makes AI flexible. It makes it better. You can get more resources. This is when AI agents need them. Windows 365 is a strong cloud platform. It hosts these smart agents. This gives them power. You get the best work. This cloud way means constant updates. Your AI stays modern. Microsoft uses its big cloud. This gives smooth connections. It gives high uptime. You get reliable AI. It works well. This also makes your platform more secure.
Future-Proofing Endpoints with Agentic Capabilities
You need to prepare your devices. This ensures they adapt. Devices will use chat AI. They will use Copilots. They will use agentic AI. This changes devices. They go from simple tools. They become active. They become smart. They become self-reliant. This means your devices guess needs. They fix problems alone. They make users happier. This keeps your Windows 365 new. It gets them ready for the future.
Securing the Agentic Future: Trust and Resilience
You must make your agentic future safe. Trust and strength are very important. Microsoft guides its AI work. It uses main rules. These rules make sure AI is good.
Microsoft's Responsible AI Principles
Microsoft builds its AI systems. They stand on strong good ideas. These rules make sure your AI agents act well.
| Responsible AI Principle | Core Meaning for Agentic Systems |
|---|---|
| Fairness | It treats every user fairly. It does not create bias. It does not make bias worse. This is based on context. It is not based on sensitive traits. |
| Reliability and Safety | It always works well. This is true in different situations. It works for unusual cases. It works for unexpected user actions. |
| Privacy and Security | It keeps user data safe. It only uses allowed information. It follows safety rules. |
| Inclusiveness | It helps many different users. This includes special needs. It includes language differences. It includes different tech skills. |
| Transparency | It clearly shows what it can do. It shows its limits. It shows how it uses data. It shows how it makes results. Users stay informed. |
| Accountability | People oversee it. It follows rules. Developers watch it. This is for its whole life. |
These rules make sure your AI systems are fair. They are also reliable. They keep things private. They help everyone.
Building Trust in Autonomous IT Systems
You need to trust IT systems that work alone. Microsoft builds this trust. It uses several ways. These ways make sure agents do what they should.
| Mechanism Dimension | Specific Control / Capability | Purpose & Trust Impact |
|---|---|---|
| Scope & Intent | It works within set limits. It sticks to its job. | It makes sure agents do only what is planned. |
| Human Control | People check high-risk actions. They approve them. They can stop them. | People still watch over things. They are responsible for big choices. |
| Observability | There are clear logs. There are feedback loops. There is checking. | You can fully see what the system does. You can see how it works. |
| Data Protection | It only accesses what it needs. It manages data life. | It stops private data from leaking. It stops too many agents. |
| Threat Mitigation | It has many layers of defense. It stops agent takeover. It stops bad inputs. | It can shut down safely. It can stand up to attacks. |
Microsoft also uses many layers of defense. This is for the model. It is for safety systems. It is for application layers. You use Microsoft's full security. You use its management system. This makes sure systems that work alone. They can be watched. They are strong. They are trusted by design.
Advanced Threat Protection for AI Workloads
Keeping your AI workloads safe from threats is key. Microsoft gives advanced safety steps. These steps protect your AI systems.
- Real-time Threat Detection: Microsoft Defender for AI and Defender for Cloud work together. They watch live. They stop AI attacks. This includes jailbreaks. It includes model poisoning. It includes prompt injections.
- Correlated Threat Intelligence: Microsoft Defender XDR links security alerts. These are from AI workloads. They link to wider company threat data. This gives you a full view. It shows possible threats.
- Graph-Based Investigations: Microsoft Sentinel gives data with graph context. This makes checking AI security problems easier.
- Automated Security Operations: Security Copilot helps Security Operations Center (SOC) teams. It uses AI. It makes finding threats faster. It makes fixing them faster.
These tools give strong security. They fight cyberattacks. They make your overall security stronger.
Ensuring Business Continuity and Recovery
You must make sure your agentic systems keep working. You must make sure they can recover. This means making your IT strong. You need plans for bad things. Early security checks help stop problems. Microsoft focuses on constant security. It focuses on security that works alone. This makes sure your systems can get better fast. It means less downtime. It protects your data. Strong Windows security. Good plans for problems. These are key for the future. They keep your security strong.
Agentic systems change IT. They make it work alone. They make it safe. They make it fast. Microsoft's AI vision adds these to Windows 365. It adds them to Intune. This makes a strong Microsoft AI system. You need to accept this change. It will happen. Learn about these new ideas. Get your company ready. AI agents will manage your digital stuff. They will keep it safe. Windows 365 and Intune will start this change.
FAQ
What are agentic systems?
Agentic systems are smart AI. They work by themselves. They learn new things. They change as needed. These agents fix hard problems. They reach goals. People do not need to tell them what to do all the time. You will see them change IT work. This is true for Windows 365.
How do Windows 365 and Intune use AI?
Windows 365 and Intune use AI. This makes management smart. AI agents do tasks by themselves. They make security better. You get smart device control. Cloud PCs work better with AI. This makes your Windows 365 experience good.
What is Windows 365 for Agents?
Windows 365 for Agents gives safe Cloud PCs. These are for your AI agents. It helps manage them fully. You can see what they do. This system helps tasks run by themselves. It makes sure your agents work well.
What benefits do Cloud PCs offer with AI?
Cloud PCs are very flexible. They give you a personal Windows. AI makes them work best. You can use them from anywhere. Windows 365 gives these strong Cloud PCs. They make your work easier. You can manage your Windows 365 easily.
How does Microsoft ensure security for AI agents?
Microsoft uses strong security. Defender XDR keeps AI safe. Purview makes sure data follows rules. You get good threat protection. This is for your Windows. This builds trust in systems that work alone. Microsoft's AI rules guide all security. This is for Windows 365.
Benefits, Use Cases, and Implementation Guidance
Key Benefits of Agentic AI in IT Operations
- Faster incident triage and root-cause analysis.
- More consistent endpoint and Cloud PC administration.
- Reduced repetitive work for service desk and operations teams.
- Earlier detection of security and compliance anomalies.
- Better scalability when infrastructure and agent workloads grow.
Common Agentic AI Use Cases
- Investigating alerts and correlating signals across tools.
- Reviewing device or policy changes before deployment.
- Automating safe onboarding and offboarding workflows.
- Optimizing Cloud PC capacity and user experience.
- Preparing remediation plans while escalating high-risk actions to people.
Implementation Checklist
- Define the business outcome and the agent's permitted scope.
- Classify actions by risk and decide where approval is mandatory.
- Give every agent a controlled identity and least-privilege access.
- Connect only approved data sources and tools.
- Log prompts, decisions, tool calls, outcomes, and overrides.
- Start read-only, measure reliability, then expand autonomy gradually.
Common Mistakes to Avoid
- Granting broad permissions before the agent has demonstrated reliability.
- Confusing a fluent response with a validated operational decision.
- Automating high-impact changes without rollback and human escalation.
- Ignoring data quality, stale inventory, and incomplete context.
- Measuring activity instead of resolution time, reliability, risk, and business value.
Pros and Cons
| Pros | Cons and Risks |
|---|---|
| Faster investigation and response | Incorrect plans can create operational impact |
| Continuous monitoring and optimization | Requires strong identity, permissions, and audit controls |
| Scalable support for hybrid environments | Integration and data-quality work can be substantial |
| Less repetitive administration | Human oversight remains essential for high-stakes actions |
Frequently Asked Questions About Agentic AI for IT Operations
What is the difference between agentic AI and generative AI?
Generative AI primarily creates or transforms content in response to a prompt. Agentic AI adds planning, tool use, execution, and feedback loops so it can pursue a goal across multiple steps under defined controls.
How does agentic AI differ from traditional automation?
Traditional automation follows predefined rules and paths. An agent can interpret context, choose among approved actions, and adapt when conditions change, but it still needs boundaries and validation.
What are the best first use cases?
Start with low-risk, repeatable tasks such as alert summarization, inventory checks, ticket classification, policy recommendations, and read-only investigations.
Can agentic AI replace IT operations engineers?
It is more likely to change their work than eliminate it. Engineers remain responsible for architecture, policy, exception handling, validation, and high-impact decisions.
How should organizations secure AI agents?
Use unique identities, least-privilege permissions, approved tools, strong authentication, data protection, continuous monitoring, and complete audit trails.
How do Windows 365 and Intune support agentic operations?
They provide managed Cloud PCs, endpoint policy, identity integration, device signals, and operational controls that agents can use within an approved governance model.
What role does Microsoft Foundry play?
Microsoft Foundry supports building and operating AI applications and agents, including model selection, connections to enterprise knowledge, and controlled deployment patterns.
What is Agent 365?
In the episode's Microsoft architecture discussion, Agent 365 is presented as a control-plane concept for discovering, governing, securing, and observing agents across the organization.
How can teams measure ROI?
Track resolution time, successful automation rate, avoided manual effort, change failure rate, security outcomes, user experience, and the cost of operating the agent.
What should remain human-controlled?
Keep high-impact identity, security, financial, compliance, and production changes subject to explicit approval until the system has proven reliable and the organization has safe rollback procedures.
What happens when an agent makes a mistake?
Use bounded permissions, approval gates, rollback, monitoring, and escalation. Every action should be attributable so operators can investigate and improve the workflow.
Is agentic AI the same as AIOps?
No. AIOps focuses on applying analytics and automation to IT operations. Agentic AI can be part of an AIOps strategy, but it adds goal-oriented planning and action capabilities.
How should an enterprise begin?
Choose one measurable, low-risk workflow; document the desired outcome; connect only necessary systems; run in read-only mode; and expand autonomy based on evidence.
Related M365.fm Resources
Continue with agentic AI architecture, the difference between agents and workflows in Copilot, Microsoft Copilot and Microsoft 365 Copilot, and Microsoft security topics.
🚀 Want to be part of m365.fm?
Then stop just listening… and start showing up.
👉 Connect with me on LinkedIn and let’s make something happen:
- 🎙️ Be a podcast guest and share your story
- 🎧 Host your own episode (yes, seriously)
- 💡 Pitch topics the community actually wants to hear
- 🌍 Build your personal brand in the Microsoft 365 space
This isn’t just a podcast — it’s a platform for people who take action.
🔥 Most people wait. The best ones don’t.
👉 Connect with me on LinkedIn and send me a message:
"I want in"
Let’s build something awesome 👊
1
00:00:00,000 --> 00:00:08,260
Yeah, welcome back to the MC65 and the podcast where we explore Microsoft AI, Cloud and point management and pack,
2
00:00:08,260 --> 00:00:10,580
while the shaping the future of enterprise IT.
3
00:00:10,580 --> 00:00:18,100
Today's guest has an extraordinary impact on Microsoft modern road place journey through the acquisition of FS Logics.
4
00:00:18,100 --> 00:00:27,180
He became the part of Microsoft and had shaped the, well, would become Azure Word will desktop and Windows 365 platform.
5
00:00:27,180 --> 00:00:32,980
Now serving organizational around the globe. Now he has changed to a new employee.
6
00:00:32,980 --> 00:00:39,100
He worked as nerve at NERDIO along the way he contribute to innovations like Microsoft 365,
7
00:00:39,100 --> 00:00:47,580
who to Windows 365 switch and those app next generation Windows AI experience, out of five books and listen.
8
00:00:47,580 --> 00:00:55,980
As an investor on multiple, yeah, the patterns covering cloud computing AI and visualization today.
9
00:00:55,980 --> 00:01:04,540
We'll diving into the evolution of cloud PCs, the rising of agentic AI and the future of into an end point management Windows 365.
10
00:01:04,540 --> 00:01:09,420
Microsoft co-pilot rather enterprise IT is having over the next decade.
11
00:01:09,420 --> 00:01:13,420
Christian, welcome to the MC65 and podcast.
12
00:01:13,420 --> 00:01:16,380
>> Yeah, Mirko, thanks to her having me.
13
00:01:16,380 --> 00:01:21,020
I love to be here. So thanks to the invite and looking forward to our conversation.
14
00:01:21,020 --> 00:01:33,820
>> Yeah, you have an extraordinary way. Can you tell a little bit how that you began and how you end up at NERDIO?
15
00:01:33,820 --> 00:01:40,220
>> Yeah, it's a good question. So first of all, I just like yourself, we were chatting before.
16
00:01:40,220 --> 00:01:48,860
We started recording about a community and a Microsoft MVP and really help each other to learn and share knowledge
17
00:01:48,860 --> 00:01:52,220
and give each other opportunities to share things you learned.
18
00:01:52,220 --> 00:02:01,740
That's really where my career started as well. So back in 2016, all 2015 pretty much like 10, 11 years ago.
19
00:02:01,740 --> 00:02:08,460
I started to embrace the community and share my knowledge and present at user groups and blog and
20
00:02:08,460 --> 00:02:15,020
all these things that's part of that great community work you do next to your day job pretty much too.
21
00:02:15,020 --> 00:02:24,380
And yeah, after that, I kind of got a lot of opportunities and at some point in 2016, 2017,
22
00:02:24,380 --> 00:02:31,500
I got an opportunity to join a company called F is Logic and F is Logic is a profile management solution
23
00:02:31,500 --> 00:02:34,700
that now is part of Microsoft as the core technology.
24
00:02:34,700 --> 00:02:42,300
I got acquired about Microsoft in 2018 and I was in a consultant doing kind of the community work on the side.
25
00:02:42,300 --> 00:02:49,740
And kind of as part of the community work, I created more opportunities kind of a little bit
26
00:02:49,740 --> 00:02:56,140
or let's say community opportunities created kind of new business opportunities for me to join
27
00:02:56,140 --> 00:03:02,380
F is Logic's and then I've got a called a choir. And that after since I've been part of Microsoft for
28
00:03:02,380 --> 00:03:08,540
eight years, A VD, Azure Virtual Desktop, the LVVD back in the days when it was still called like that.
29
00:03:08,540 --> 00:03:16,140
And then eventually in 2020, I moved from the A VD to a project called Project Your Shoots which
30
00:03:16,140 --> 00:03:22,860
turned out to be Winnus 365. So building an installed mode for one and a half years. First in the
31
00:03:22,860 --> 00:03:29,100
Netherlands for nine months and then I moved with my family to Reppman's so I worked, I lived there
32
00:03:29,100 --> 00:03:36,780
for five years. We just came back as that product when Skyrocket with COVID remote working was more
33
00:03:36,780 --> 00:03:43,740
popular than I before. I kind of worked on a couple kind of example scenarios and features you
34
00:03:43,740 --> 00:03:51,820
just called out. But at some point, I really saw kind of the partner landscape evolving AI and
35
00:03:51,820 --> 00:04:00,540
what you said like entry 65, co-pilot, like agent AI. So I saw an opportunity to go work for the
36
00:04:00,540 --> 00:04:06,940
largest partner of Winnus Cloud, so Winnus 365 and actually worked as a which is NERDIO, to build a whole new
37
00:04:06,940 --> 00:04:14,940
platform from scratch and really hands and embrace the Winnus 365 and in tune story. And that's the
38
00:04:14,940 --> 00:04:19,980
platform that I'm currently building and I'm spearheading that team with like roughly 20 people there
39
00:04:19,980 --> 00:04:25,980
and it's really excited to see as the, yeah, the first start of that kind of set of experiences is
40
00:04:25,980 --> 00:04:32,140
coming to life and an age two, this kind of the year we will have more releases there to come but it
41
00:04:32,140 --> 00:04:38,860
all starts from yeah community and people and and we're back from the customer to a better product
42
00:04:38,860 --> 00:04:46,780
because I honestly believe that a product is the representation of the people and the culture who
43
00:04:46,780 --> 00:04:52,540
build it. And yeah, if your product is great then you most likely have a great team that works on
44
00:04:52,540 --> 00:04:59,500
it as well and if your product lacks experiences or suffers and sometimes it's yeah something different
45
00:04:59,500 --> 00:05:06,780
so. So kind of my journey, so I've been in Microsoft for like eight years I sat, I've been
46
00:05:06,780 --> 00:05:13,260
shipping a lot of features, probably a lot of patents, I have a couple of those sitting here that
47
00:05:13,260 --> 00:05:19,580
I brought with me here in my, all museum and my home office here and we can talk a little bit about
48
00:05:19,580 --> 00:05:27,340
it but like fast forward like I think what is important to kind of realize here is that, yeah,
49
00:05:27,340 --> 00:05:33,740
windows and operating systems have been kind of the same for like 20, 30 years. What you're seeing
50
00:05:33,740 --> 00:05:40,220
shifting right now is like in the age of AI, UI and the way how we interact with computers is changing
51
00:05:40,220 --> 00:05:47,180
and cloud PCs are more and more important because they run as a windows PC in the cloud and you can
52
00:05:47,180 --> 00:05:52,700
travel or connect to it everywhere you are without like the complexity of hardware burden or
53
00:05:52,700 --> 00:05:59,340
limitations of of hardware these days where RAM and other hardware is kind of increasing
54
00:05:59,340 --> 00:06:05,020
in terms of price points that that's an exciting area we're in right now and I guess we are going
55
00:06:05,020 --> 00:06:09,660
to I'm going to talk a little bit about that too right, Mikra? Yeah, awesome. So yeah,
56
00:06:09,660 --> 00:06:17,340
then there was my first question, I see, or how Azure Virtual Desktop and Windows 365
57
00:06:17,340 --> 00:06:25,180
Grunt are complementary? Yeah, it's a good question. So I guess that's probably the one most kind
58
00:06:25,180 --> 00:06:32,380
of a real current question that I'm getting still since the release of Windows 365 still till today.
59
00:06:32,380 --> 00:06:38,540
So Azure Virtual Desktop is a platform as a source of solutions of passing short, I think we're all
60
00:06:38,540 --> 00:06:45,500
familiar with past and past these days. So it's kind of the I always compare it a little bit with the
61
00:06:45,500 --> 00:06:54,540
the pizza kind of menu slash who begs the pizza and who is serving the pizza completely out of
62
00:06:54,540 --> 00:07:00,060
the out of the oven to you on a plate. So that's kind of the difference between Azure Virtual Desktop
63
00:07:00,060 --> 00:07:05,980
and Windows 365. Azure Virtual Desktop you have all the buttons and things you can do with a
64
00:07:05,980 --> 00:07:10,860
traditional VDI solution but then with the control plate and hosted as a pass offering
65
00:07:10,860 --> 00:07:17,420
and then a Windows 365 is really the SAS offering where to VM or to cloud PC and everything
66
00:07:17,420 --> 00:07:22,700
on the back, scalability, image management, all the things you normally have to do yourself in an
67
00:07:22,700 --> 00:07:31,500
IT department is managed for you as part of the NASA lay, which makes it like kind of a democratizing
68
00:07:31,500 --> 00:07:37,420
way of using VDI. So it's a me and you probably are not a VDI expert. I share like more like
69
00:07:37,420 --> 00:07:46,540
M365 and all these things right and like imagine that you can all of a sudden use the VDI technology that
70
00:07:46,540 --> 00:07:52,940
was like very specialized knowledge for a small group of people that could do it. So now all of a
71
00:07:52,940 --> 00:07:59,340
sudden everyone that has M365 and inch of knowledge, you can spend up a cloud PC, manage and maintain
72
00:07:59,340 --> 00:08:07,020
it the same way as a physical PC and then make that work for your business for a very streamlined
73
00:08:07,020 --> 00:08:13,740
way and consolidate the way with the Fender and other services, co-piglet agents all coming to get
74
00:08:13,740 --> 00:08:19,500
as one. And I think that's the real difference between the two. However, Azure Virtual Desktop can be
75
00:08:19,500 --> 00:08:25,900
if you set a power scaling, outer scaling, can be a little cheaper in terms of the price but then
76
00:08:25,900 --> 00:08:31,500
there's a certain overhead that you need to fulfill to manage and maintain that yourself as an
77
00:08:31,500 --> 00:08:38,220
IT department as a partner or a solution like nerdy or where you can help and do it all together
78
00:08:38,220 --> 00:08:45,180
for off-products in one single pane of glass. But that's kind of the beauty of this ecosystem. You can
79
00:08:45,180 --> 00:08:51,260
pick and choose with this best your needs and kind of have something that yeah, it kind of
80
00:08:51,820 --> 00:08:58,220
sits in the in the direction you want to be today and tomorrow and decide where you want to be and
81
00:08:58,220 --> 00:09:03,420
yeah, that's a little bit of a long ass on a short question but I think yeah,
82
00:09:03,420 --> 00:09:10,060
Pass and SaaS is probably the best differentiation of how you define your future and how you should decide
83
00:09:10,060 --> 00:09:18,380
which product fits best or combine them either way to make it fit and use AVD and Winner 365 together.
84
00:09:19,260 --> 00:09:26,300
Yeah, you have see AVD for a long time. How has it evolved over the, yeah, from the early versions to now?
85
00:09:26,300 --> 00:09:35,580
Yeah, it's a good question. I still remember back in 2016, I think it was, I was in Redmond and I think
86
00:09:35,580 --> 00:09:44,380
it was Studio G, one of the buildings in the Redmond campus and we got a sneak peek into what was
87
00:09:44,380 --> 00:09:51,740
called Project RDMI or so remote desktop kind of, modern infrastructure was called. There was
88
00:09:51,740 --> 00:09:56,460
some motor project names before that but that was pretty much the well-known one. And that was the
89
00:09:56,460 --> 00:10:04,540
project name for modernizing RDS into a more kind of past kind of fashion cloud like Azure was just
90
00:10:04,540 --> 00:10:13,340
coming up and everyone in 2016 was looking at Azure ASM and an ARM like the eBTS supporter, you probably
91
00:10:13,340 --> 00:10:19,900
remembered it as well when the old school blew as your portal was still there. And before that we even had
92
00:10:19,900 --> 00:10:25,660
like an other service called Azure Remote App, that one didn't shut well but that was a large set
93
00:10:25,660 --> 00:10:31,900
of learning experiences, the engineering gain to eventually build Azure Virtua desktop or WVD which
94
00:10:31,900 --> 00:10:40,940
was RDMI and that evolved from being a very power shell, heavy kind of setup meaning a bunch of
95
00:10:40,940 --> 00:10:47,500
rolls, power shell scripts and hook up connections to Azure and then spins up something and it was
96
00:10:47,500 --> 00:10:53,740
Azure Marketplace kind of a Roman full but only bit like post-pool creation but didn't do the
97
00:10:53,740 --> 00:11:01,100
initial setup. So it was like let's say a very complex enrollment process at the time and then
98
00:11:01,100 --> 00:11:10,860
yeah we kind of entered the COVID area in like 2019 and it's kind of if you look back it's
99
00:11:10,860 --> 00:11:16,460
a very negative kind of time in our lives like it's like almost like few three years swapped out of
100
00:11:16,460 --> 00:11:22,860
your life and but on the on the positive side from a business perspective or career perspective I'm
101
00:11:22,860 --> 00:11:29,100
gonna say products for remote working were popular than ever before because yeah you were sent back
102
00:11:29,100 --> 00:11:34,780
home and your computer was still at the at the office though how are you the appropriate
103
00:11:34,780 --> 00:11:42,860
applications so all of a sudden Azure Virtua desktop was the most popular Azure service at Microsoft
104
00:11:42,860 --> 00:11:50,060
to yeah enable you for remote working so at that time I joined Microsoft and yeah you can imagine
105
00:11:50,060 --> 00:11:57,420
that people on the team and myself were like parachute it into companies to help onboarding them
106
00:11:57,420 --> 00:12:04,940
to Azure Virtua desktop and that happened in a like rapid pace manner that went from like let's say
107
00:12:04,940 --> 00:12:14,140
100 customers using the service in 2017 as to like thousands of thousands of customers in a year later
108
00:12:14,140 --> 00:12:19,900
when COVID happened and now I cannot disclose the number but it's like it's a very high number using
109
00:12:19,900 --> 00:12:26,380
the service and then when is 365 happened exactly at the time that more and more simplification was
110
00:12:26,380 --> 00:12:32,860
needed and SaaS to kind of build on top of that demand for more customers so yeah it's
111
00:12:32,860 --> 00:12:39,340
it has been an amazing journey to see and demand and feedback from customers if all the service to
112
00:12:39,340 --> 00:12:47,340
how good it is right now in Azure and in in in June yeah what what would you say are the biggest
113
00:12:47,340 --> 00:12:56,060
technical challenges from AVD so Microsoft say okay we built something new we built Windows
114
00:12:56,060 --> 00:13:05,100
365 yeah it's I think I think it really depends on what type of customer you are so if you're a
115
00:13:05,100 --> 00:13:11,660
customer and you have been using virtualization for the last 10 20 years primarily Citrix or
116
00:13:11,660 --> 00:13:18,220
VMware or like Omnisah or Ryzen how it's called right now or maybe Amazon Workspace is there
117
00:13:18,220 --> 00:13:23,500
various of these virtualization services out there and if you're used to that then you have that
118
00:13:23,500 --> 00:13:29,580
knowledge and expertise of virtualization in your IT department so you know kind of what to do and
119
00:13:29,580 --> 00:13:36,460
that the incremental delta changes that you need to make in order to to relearn a little bit the
120
00:13:36,460 --> 00:13:42,140
platform but the basics of virtualization remain to be the same then Azure Virtual Desktop is a
121
00:13:42,140 --> 00:13:49,260
very logical pick for modernizing from on-premises to like let's say Azure but we also see a lot of
122
00:13:49,260 --> 00:13:56,780
customers going straight from traditional virtualization to Winnus 365 and then kind of work backwards
123
00:13:56,780 --> 00:14:02,460
to Azure Virtual Desktop so they kind of use it in a hybrid way but they shoot for Winnus 365 to go
124
00:14:02,460 --> 00:14:10,620
straight for modernization to anything in June I'm 365 and then if it doesn't work like you have
125
00:14:10,620 --> 00:14:14,940
these PCs you probably have been doing them as well and then you pick a couple of these like
126
00:14:16,140 --> 00:14:20,700
applications that like stress tests the rest of your environment and if it goes true then
127
00:14:20,700 --> 00:14:28,060
let's go production and we still see a lot of customers struggling sometimes but hey I have a
128
00:14:28,060 --> 00:14:34,460
a database or an internet or some some data sovereignty like things need to just run in my own
129
00:14:34,460 --> 00:14:41,500
data center close back to the machines or back-end or there's a server running there and we cannot move
130
00:14:41,500 --> 00:14:47,660
or migrate it so and there's a solution for that now too called Azure Virtual Desktop Hybrid
131
00:14:47,660 --> 00:14:54,700
so AVD Hyde in short and then it allows you to have AVD as a control plane so that pass off-ring
132
00:14:54,700 --> 00:15:02,620
but still connect to your own on-prem data center if it runs like Nutanix or the Broadcom VM
133
00:15:02,620 --> 00:15:12,620
where ESX solution or IPv so there are new ways now to kind of mix and match AVD on-prem, AVD in the
134
00:15:12,620 --> 00:15:19,260
cloud or Winnus 365 and kind of overarge that altogether as a whole is something I think is
135
00:15:19,260 --> 00:15:26,380
his very important these days because imagine your customer and you kind of on that journey and you
136
00:15:26,380 --> 00:15:33,020
have AI or you have AVD here Winnus 365 there and all these solutions come together can be pretty
137
00:15:33,020 --> 00:15:39,020
complex to kind of navigate and know hey what's the best pick for my business so combining them in a
138
00:15:39,020 --> 00:15:45,180
good way and making sure you make these decisions while not upfront kind of going in a direction you
139
00:15:45,180 --> 00:15:51,900
don't want to so kind of paved the way towards the future without making decisions to go in a direction
140
00:15:51,900 --> 00:16:00,220
that I don't know let's say you decided to build a new house and you set the foundation for your
141
00:16:00,220 --> 00:16:05,740
house and you make a decision on the foundation that in five years or so you need to kind of
142
00:16:05,740 --> 00:16:10,540
completely rep out the foundation and rebuild your house you don't want to do that you want to kind
143
00:16:10,540 --> 00:16:18,220
of build the foundation in a way that the new parts of your house will be sustainable future and
144
00:16:18,220 --> 00:16:24,540
I think that's where like solutions like a Nordeo kick in that really overlays that and do the things
145
00:16:24,540 --> 00:16:31,420
that you can kind of simplify as a solution for you to not make the decision faster but it still
146
00:16:31,420 --> 00:16:38,300
hooks into the native solution so you're not like let's say you're not making a decision like
147
00:16:38,300 --> 00:16:44,460
you open a door and you kind of go back out so it's like easy to to leverage and simplify and as
148
00:16:44,460 --> 00:16:52,540
well to yeah kind of in the future go completely native you can do that too easily without removing
149
00:16:52,540 --> 00:16:57,580
your completely in a fire and I think that's the key to success as well as where you want to be
150
00:16:57,580 --> 00:17:04,460
making your decisions on like that you yeah bet on a on a future that is sustainable and
151
00:17:04,460 --> 00:17:11,420
guide you to and I think that's important the more complexity as well is what are kind of what's
152
00:17:11,420 --> 00:17:18,460
being added right now with AI it's it's more and more important there to yeah kind of pave that
153
00:17:18,460 --> 00:17:26,060
way towards a unified story because AI is great and AI can do a lot of things it can
154
00:17:26,060 --> 00:17:31,100
from summarizing creating docs to automating your calendar or to
155
00:17:31,100 --> 00:17:38,380
build a whole code base that normally takes months to build but there are so many models so many
156
00:17:38,380 --> 00:17:45,340
solutions out there that it kind of asks for the need of more governance and orchestration in that
157
00:17:45,340 --> 00:17:52,460
space too so I feel I'm frustrated that's becoming more and more important that all these solutions are
158
00:17:52,460 --> 00:17:59,180
kind of kind of kind of coming together as one but like from different layers as well so
159
00:17:59,180 --> 00:18:04,940
so yeah I'm curious would you think Daryl Michael what would you be in your in your space where
160
00:18:04,940 --> 00:18:14,220
you're active then yeah I see that virtualization becomes yeah in the last 20 years more and more
161
00:18:14,220 --> 00:18:22,620
important yeah I think when you're an admin I think the most people yeah don't have anything to do
162
00:18:22,620 --> 00:18:34,380
with this topic remotely so yeah I see it's it's really interesting what lessons from virtualization
163
00:18:34,380 --> 00:18:43,100
still apply today while you think the basics pretty much so so you still have a protocol you still have
164
00:18:43,100 --> 00:18:49,660
a client you still have latency and bandwidth and throughput and yeah download and upload speed
165
00:18:49,660 --> 00:18:56,300
that you need to take care of but the good thing is that the protocols have been commoditized
166
00:18:56,300 --> 00:19:03,180
pretty much and so like let's say 20 years ago there was always like you had the ctrux HDX and then
167
00:19:03,180 --> 00:19:09,260
you had the VM your blast protocols and then this one is better than the other and I'm always at
168
00:19:09,260 --> 00:19:15,100
quite you probably remember that time the even colored protocol wash or something at some time
169
00:19:15,100 --> 00:19:20,380
that time is over I think there might be still some slight differences but nobody knows is that
170
00:19:20,380 --> 00:19:26,220
anymore it's like commodity it's like it's like streaming like from your best I don't know you have
171
00:19:26,220 --> 00:19:30,460
Amazon prime you have Netflix you have all these streaming engines and they all stream with a
172
00:19:30,460 --> 00:19:36,940
click of a button and you expect it to work that's kind of what protocols kind of work these days too
173
00:19:36,940 --> 00:19:42,140
like if you stream something if you have like the speed upload and download speed of what can
174
00:19:42,140 --> 00:19:47,820
stream in that like series you can stream like virtualization these days and that's the beauty
175
00:19:47,820 --> 00:19:57,020
of a SaaS solution is that at whole kind of let's say TCP UDP stun turn that kind of optimizing the
176
00:19:57,020 --> 00:20:03,340
the streaming engine like that normally requires a lot of knowledge and I don't know
177
00:20:03,340 --> 00:20:12,700
napscaler or firewalling settings and that's all gone so so I think that's that's one of the things
178
00:20:12,700 --> 00:20:17,740
that you need to kind of have a little bit of understanding of like what it is because it helps with
179
00:20:17,740 --> 00:20:23,980
the troubleshooting if you're like having network issues or let's say you're I don't know your
180
00:20:23,980 --> 00:20:31,100
company decided to change from ISPs or there's a different set of firewall rules coming true or
181
00:20:31,100 --> 00:20:36,220
VPN tunneling is reconfigured that you need to kind of have a little bit of knowledge there how to
182
00:20:36,220 --> 00:20:43,580
routing and re-riring and that flow works but for the rest I think yeah clients are still the same
183
00:20:43,580 --> 00:20:49,580
from the fact that you need a client in order to have the redirections and the streaming happening
184
00:20:49,580 --> 00:20:55,660
and the windows client like the windows app for example is still the best client out there with the
185
00:20:55,660 --> 00:21:01,900
most feature rich capabilities because of course Microsoft owns windows and that's the code base
186
00:21:01,900 --> 00:21:08,140
my phone so that kind of is naturally always the best solution but it doesn't mean that
187
00:21:08,140 --> 00:21:16,060
that the solution for windows and Microsoft is the only one that works great like Mac and all
188
00:21:16,060 --> 00:21:23,420
these other platforms like even Android has a windows app now that feels agnosticly the same as the
189
00:21:23,420 --> 00:21:30,780
other platforms and have teams optimizations and folder redirection disc redirection,
190
00:21:30,780 --> 00:21:35,340
clibboard redirection, condition access one way, two way all these capabilities are now in there
191
00:21:35,340 --> 00:21:43,180
cross-plat per client which originally was only like windows or weps so I think a great enhancement
192
00:21:43,180 --> 00:21:50,220
that we have seen over time when the windows 365 and I should first start to come out there
193
00:21:50,220 --> 00:21:56,380
okay miko where are you
194
00:21:56,380 --> 00:21:56,940
here?
195
00:21:57,580 --> 00:21:57,580
you
196
00:21:57,580 --> 00:22:03,380
[ Silence ]
197
00:22:04,380 --> 00:22:13,380
[ Silence ]
198
00:22:13,380 --> 00:22:23,380
[ Silence ]
199
00:22:24,380 --> 00:22:33,380
[ Silence ]
200
00:22:33,380 --> 00:22:43,380
[ Silence ]
201
00:22:44,380 --> 00:22:53,380
[ Silence ]
202
00:22:53,380 --> 00:23:13,380
[ Silence ]
203
00:23:13,380 --> 00:23:17,380
-Negro, are you back?
204
00:23:17,380 --> 00:23:18,380
-I'm back.
205
00:23:18,380 --> 00:23:21,380
-I was kicked out.
206
00:23:21,380 --> 00:23:24,380
I don't know what's happened.
207
00:23:24,380 --> 00:23:27,380
-We were talking about streaming and incidental,
208
00:23:27,380 --> 00:23:29,380
and download, and listen.
209
00:23:29,380 --> 00:23:35,380
-That is the fun part.
210
00:23:35,380 --> 00:23:40,380
Damn, I have to look at this as records.
211
00:23:40,380 --> 00:23:44,380
-It's still recording here, so I think we're good.
212
00:23:44,380 --> 00:23:50,380
I just kept going, so you can just after that cut it out and stitch it.
213
00:23:50,380 --> 00:23:54,380
-Yeah, sorry.
214
00:23:54,380 --> 00:23:56,380
-I'm using it.
215
00:23:56,380 --> 00:24:00,380
-I can just go with the flow.
216
00:24:00,380 --> 00:24:01,380
-Yeah.
217
00:24:01,380 --> 00:24:05,380
-I think that's really interesting,
218
00:24:05,380 --> 00:24:09,380
but you're expert in Windows 365,
219
00:24:09,380 --> 00:24:12,380
and there are a lot of buzzwords.
220
00:24:12,380 --> 00:24:14,380
There are two.
221
00:24:14,380 --> 00:24:17,380
What is Windows 365 boot?
222
00:24:17,380 --> 00:24:21,380
-What is Windows 365?
223
00:24:21,380 --> 00:24:25,380
-Yes, it's a fun story.
224
00:24:25,380 --> 00:24:28,380
Originally, Windows 365 boot,
225
00:24:28,380 --> 00:24:31,380
so there was a person, part of Microsoft,
226
00:24:31,380 --> 00:24:33,380
called Melissa Grand.
227
00:24:33,380 --> 00:24:35,380
I think she works for Amazon now.
228
00:24:35,380 --> 00:24:39,380
She was the marketing director for Windows 365,
229
00:24:39,380 --> 00:24:43,380
and we were brainstorming about a feature, like, let's say,
230
00:24:43,380 --> 00:24:46,380
I think it was three, four years ago, four years ago.
231
00:24:46,380 --> 00:24:52,380
I think right now, to kind of solve the challenge of virtualization
232
00:24:52,380 --> 00:24:53,380
for end users,
233
00:24:53,380 --> 00:24:59,380
where, like imagine you're, like, walking down to your office,
234
00:24:59,380 --> 00:25:02,380
and you need to explain to just a normal user
235
00:25:02,380 --> 00:25:05,380
that isn't a technical savvy user,
236
00:25:05,380 --> 00:25:08,380
to log onto a VDI environment.
237
00:25:08,380 --> 00:25:10,380
So that's very hard.
238
00:25:10,380 --> 00:25:15,380
It's almost requiring a PDF, a couple calls,
239
00:25:15,380 --> 00:25:19,380
an upset user, because a couple of tanks don't work,
240
00:25:19,380 --> 00:25:21,380
and then all of a sudden, you're a couple hours later,
241
00:25:21,380 --> 00:25:26,380
and the user is on their first day ramping up into the new system,
242
00:25:26,380 --> 00:25:28,380
and it's not satisfied.
243
00:25:28,380 --> 00:25:30,380
So we were like, thinking, how can we streamline that?
244
00:25:30,380 --> 00:25:32,380
From, hey, you have a PC,
245
00:25:32,380 --> 00:25:34,380
and then you need to install a client,
246
00:25:34,380 --> 00:25:36,380
and then in the client, you feel as a user,
247
00:25:36,380 --> 00:25:40,380
hey, you go from Windows to another Windows solar in the cloud,
248
00:25:40,380 --> 00:25:42,380
like, how does that work?
249
00:25:42,380 --> 00:25:45,380
And then the user, you know, you need to use the Windows
250
00:25:45,380 --> 00:25:47,380
to use the Windows computer,
251
00:25:47,380 --> 00:25:49,380
and then you can use the Windows computer,
252
00:25:49,380 --> 00:25:51,380
and then you can, you know,
253
00:25:51,380 --> 00:25:52,380
you need to use the Windows computer,
254
00:25:52,380 --> 00:25:55,380
and then you can use the Windows computer,
255
00:25:55,380 --> 00:25:57,380
and then you can just use the Windows computer,
256
00:25:57,380 --> 00:25:59,380
and then you, you know,
257
00:25:59,380 --> 00:26:01,380
you can use the Windows computer,
258
00:26:01,380 --> 00:26:03,380
and then you just, you know,
259
00:26:03,380 --> 00:26:05,380
you can use the Windows computer,
260
00:26:05,380 --> 00:26:07,380
you can use the Windows computer,
261
00:26:07,380 --> 00:26:09,380
and then you can use the Windows computer,
262
00:26:09,380 --> 00:26:11,380
and then you need to, you know,
263
00:26:11,380 --> 00:26:13,380
and then you need to use the Windows computer,
264
00:26:13,380 --> 00:26:15,380
and then you need to use the Windows computer,
265
00:26:15,380 --> 00:26:17,380
and then you need to use the Windows computer,
266
00:26:17,380 --> 00:26:19,380
and then you need to use the Windows computer,
267
00:26:19,380 --> 00:26:21,380
and then you need to use the Windows computer,
268
00:26:21,380 --> 00:26:23,380
and then you need to use the Windows computer,
269
00:26:23,380 --> 00:26:25,380
and then you need to use the Windows computer,
270
00:26:25,380 --> 00:26:27,380
and then you need to use the Windows computer,
271
00:26:27,380 --> 00:26:29,380
and then you need to use the Windows computer,
272
00:26:29,380 --> 00:26:31,380
and then you need to use the Windows computer,
273
00:26:31,380 --> 00:26:33,380
and then you need to use the Windows computer,
274
00:26:33,380 --> 00:26:35,380
and then you need to use the Windows computer,
275
00:26:35,380 --> 00:26:37,380
and then you need to use the Windows computer,
276
00:26:37,380 --> 00:26:39,380
and then you need to use the Windows computer,
277
00:26:39,380 --> 00:26:41,380
and then you need to use the Windows computer,
278
00:26:41,380 --> 00:26:43,380
and then you need to use the Windows computer,
279
00:26:43,380 --> 00:26:45,380
and then you need to use the Windows computer,
280
00:26:45,380 --> 00:26:47,380
and then you need to use the Windows computer,
281
00:26:47,380 --> 00:26:49,380
and then you need to use the Windows computer,
282
00:26:49,380 --> 00:26:51,380
and then you need to use the Windows computer,
283
00:26:51,380 --> 00:26:53,380
and then you need to use the Windows computer,
284
00:26:53,380 --> 00:26:55,380
and then you need to use the Windows computer,
285
00:26:55,380 --> 00:26:57,380
and then you need to use the Windows computer,
286
00:26:57,380 --> 00:26:59,380
and then you need to use the Windows computer,
287
00:26:59,380 --> 00:27:01,380
and then you need to use the Windows computer,
288
00:27:01,380 --> 00:27:03,380
and then you need to use the Windows computer,
289
00:27:03,380 --> 00:27:05,380
and then you need to use the Windows computer,
290
00:27:05,380 --> 00:27:07,380
and then you need to use the Windows computer,
291
00:27:07,380 --> 00:27:09,380
and then you need to use the Windows computer,
292
00:27:09,380 --> 00:27:11,380
and then you need to use the Windows computer,
293
00:27:11,380 --> 00:27:13,380
and then you need to use the Windows computer,
294
00:27:13,380 --> 00:27:15,380
and then you need to use the Windows computer,
295
00:27:15,380 --> 00:27:17,380
and then you need to use the Windows computer,
296
00:27:17,380 --> 00:27:19,380
and then you need to use the Windows computer,
297
00:27:19,380 --> 00:27:21,380
and then you need to use the Windows computer,
298
00:27:21,380 --> 00:27:23,380
and then you need to use the Windows computer,
299
00:27:23,380 --> 00:27:25,380
and then you need to use the Windows computer,
300
00:27:25,380 --> 00:27:27,380
and then you need to use the Windows computer,
301
00:27:27,380 --> 00:27:29,380
and then you need to use the Windows computer,
302
00:27:29,380 --> 00:27:31,380
and then you need to use the Windows computer,
303
00:27:31,380 --> 00:27:33,380
and then you need to use the Windows computer,
304
00:27:33,380 --> 00:27:35,380
and then you need to use the Windows computer,
305
00:27:35,380 --> 00:27:37,380
and then you need to use the Windows computer,
306
00:27:37,380 --> 00:27:39,380
and then you need to use the Windows computer,
307
00:27:39,380 --> 00:27:41,380
and then you need to use the Windows computer,
308
00:27:41,380 --> 00:27:43,380
and then you need to use the Windows computer,
309
00:27:43,380 --> 00:27:45,380
and then you need to use the Windows computer,
310
00:27:45,380 --> 00:27:47,380
and then you need to use the Windows computer,
311
00:27:47,380 --> 00:27:49,380
and then you need to use the Windows computer,
312
00:27:49,380 --> 00:27:51,380
and then you need to use the Windows computer,
313
00:27:51,380 --> 00:27:53,380
and then you need to use the Windows computer,
314
00:27:53,380 --> 00:27:55,380
and then you need to use the Windows computer,
315
00:27:55,380 --> 00:27:57,380
and then you need to use the Windows computer,
316
00:27:57,380 --> 00:27:59,380
and then you need to use the Windows computer,
317
00:27:59,380 --> 00:28:01,380
and then you need to use the Windows computer,
318
00:28:01,380 --> 00:28:03,380
and then you need to use the Windows computer,
319
00:28:03,380 --> 00:28:05,380
and then you need to use the Windows computer,
320
00:28:05,380 --> 00:28:07,380
and then you need to use the Windows computer,
321
00:28:07,380 --> 00:28:09,380
and then you need to use the Windows computer,
322
00:28:09,380 --> 00:28:11,380
and then you need to use the Windows computer,
323
00:28:11,380 --> 00:28:13,380
and then you need to use the Windows computer,
324
00:28:13,380 --> 00:28:15,380
and then you need to use the Windows computer,
325
00:28:15,380 --> 00:28:17,380
and then you need to use the Windows computer,
326
00:28:17,380 --> 00:28:19,380
and then you need to use the Windows computer,
327
00:28:19,380 --> 00:28:21,380
and then you need to use the Windows computer,
328
00:28:21,380 --> 00:28:23,380
and then you need to use the Windows computer,
329
00:28:23,380 --> 00:28:25,380
and then you need to use the Windows computer,
330
00:28:25,380 --> 00:28:27,380
and then you need to use the Windows computer,
331
00:28:27,380 --> 00:28:29,380
and then you need to use the Windows computer,
332
00:28:29,380 --> 00:28:31,380
and then you need to use the Windows computer,
333
00:28:31,380 --> 00:28:33,380
and then you need to use the Windows computer,
334
00:28:33,380 --> 00:28:35,380
and then you need to use the Windows computer,
335
00:28:35,380 --> 00:28:37,380
and then you need to use the Windows computer,
336
00:28:37,380 --> 00:28:39,380
and then you need to use the Windows computer,
337
00:28:39,380 --> 00:28:41,380
and then you need to use the Windows computer,
338
00:28:41,380 --> 00:28:43,380
and then you need to use the Windows computer,
339
00:28:43,380 --> 00:28:45,380
and then you need to use the Windows computer,
340
00:28:45,380 --> 00:28:47,380
and then you need to use the Windows computer,
341
00:28:47,380 --> 00:28:49,380
and then you need to use the Windows computer,
342
00:28:49,380 --> 00:28:51,380
and then you need to use the Windows computer,
343
00:28:51,380 --> 00:28:53,380
and then you need to use the Windows computer,
344
00:28:53,380 --> 00:28:55,380
and then you need to use the Windows computer,
345
00:28:55,380 --> 00:28:57,380
and then you need to use the Windows computer,
346
00:28:57,380 --> 00:28:59,380
and then you need to use the Windows computer,
347
00:28:59,380 --> 00:29:01,380
and then you need to use the Windows computer,
348
00:29:01,380 --> 00:29:03,380
and then you need to use the Windows computer,
349
00:29:03,380 --> 00:29:05,380
and then you need to use the Windows computer,
350
00:29:05,380 --> 00:29:07,380
and then you need to use the Windows computer,
351
00:29:07,380 --> 00:29:09,380
and then you need to use the Windows computer,
352
00:29:09,380 --> 00:29:11,380
and then you need to use the Windows computer,
353
00:29:11,380 --> 00:29:13,380
and then you need to use the Windows computer,
354
00:29:13,380 --> 00:29:15,380
and then you need to use the Windows computer,
355
00:29:15,380 --> 00:29:17,380
and then you need to use the Windows computer,
356
00:29:17,380 --> 00:29:19,380
and then you need to use the Windows computer,
357
00:29:19,380 --> 00:29:21,380
and then you need to use the Windows computer,
358
00:29:21,380 --> 00:29:23,380
and then you need to use the Windows computer,
359
00:29:23,380 --> 00:29:25,380
and then you need to use the Windows computer,
360
00:29:25,380 --> 00:29:27,380
and then you need to use the Windows computer,
361
00:29:27,380 --> 00:29:29,380
and then you need to use the Windows computer,
362
00:29:29,380 --> 00:29:31,380
and then you need to use the Windows computer,
363
00:29:31,380 --> 00:29:33,380
and then you need to use the Windows computer,
364
00:29:33,380 --> 00:29:35,380
and then you need to use the Windows computer,
365
00:29:35,380 --> 00:29:37,380
and then you need to use the Windows computer,
366
00:29:37,380 --> 00:29:39,380
and then you need to use the Windows computer,
367
00:29:39,380 --> 00:29:41,380
and then you need to use the Windows computer,
368
00:29:41,380 --> 00:29:43,380
and then you need to use the Windows computer,
369
00:29:43,380 --> 00:29:45,380
and then you need to use the Windows computer,
370
00:29:45,380 --> 00:29:47,380
and then you need to use the Windows computer,
371
00:29:47,380 --> 00:29:49,380
and then you need to use the Windows computer,
372
00:29:49,380 --> 00:29:51,380
and then you need to use the Windows computer,
373
00:29:51,380 --> 00:29:53,380
and then you need to use the Windows computer,
374
00:29:53,380 --> 00:29:55,380
and then you need to use the Windows computer,
375
00:29:55,380 --> 00:29:57,380
and then you need to use the Windows computer,
376
00:29:57,380 --> 00:29:59,380
and then you need to use the Windows computer,
377
00:29:59,380 --> 00:30:01,380
and then you need to use the Windows computer,
378
00:30:01,380 --> 00:30:03,380
and then you need to use the Windows computer,
379
00:30:03,380 --> 00:30:05,380
and then you need to use the Windows computer,
380
00:30:05,380 --> 00:30:07,380
and then you need to use the Windows computer,
381
00:30:07,380 --> 00:30:09,380
and then you need to use the Windows computer,
382
00:30:09,380 --> 00:30:11,380
and then you need to use the Windows computer,
383
00:30:11,380 --> 00:30:13,380
and then you need to use the Windows computer,
384
00:30:13,380 --> 00:30:15,380
and then you need to use the Windows computer,
385
00:30:15,380 --> 00:30:17,380
and then you need to use the Windows computer,
386
00:30:17,380 --> 00:30:19,380
and then you need to use the Windows computer,
387
00:30:19,380 --> 00:30:21,380
and then you need to use the Windows computer,
388
00:30:21,380 --> 00:30:23,380
and then you need to use the Windows computer,
389
00:30:23,380 --> 00:30:25,380
and then you need to use the Windows computer,
390
00:30:25,380 --> 00:30:27,380
and then you need to use the Windows computer,
391
00:30:27,380 --> 00:30:29,380
and then you need to use the Windows computer,
392
00:30:29,380 --> 00:30:31,380
and then you need to use the Windows computer,
393
00:30:31,380 --> 00:30:33,380
and then you need to use the Windows computer,
394
00:30:33,380 --> 00:30:35,380
and then you need to use the Windows computer,
395
00:30:35,380 --> 00:30:37,380
and then you need to use the Windows computer,
396
00:30:37,380 --> 00:30:39,380
and then you need to use the Windows computer,
397
00:30:39,380 --> 00:30:41,380
and then you need to use the Windows computer,
398
00:30:41,380 --> 00:30:43,380
and then you need to use the Windows computer,
399
00:30:43,380 --> 00:30:45,380
and then you need to use the Windows computer,
400
00:30:45,380 --> 00:30:47,380
and then you need to use the Windows computer,
401
00:30:47,380 --> 00:30:49,380
and then you need to use the Windows computer,
402
00:30:49,380 --> 00:30:51,380
and then you need to use the Windows computer,
403
00:30:51,380 --> 00:30:53,380
and then you need to use the Windows computer,
404
00:30:53,380 --> 00:30:55,380
and then you need to use the Windows computer,
405
00:30:55,380 --> 00:30:57,380
and then you need to use the Windows computer,
406
00:30:57,380 --> 00:30:59,380
and then you need to use the Windows computer,
407
00:30:59,380 --> 00:31:01,380
and then you need to use the Windows computer,
408
00:31:01,380 --> 00:31:03,380
and then you need to use the Windows computer,
409
00:31:03,380 --> 00:31:05,380
and then you need to use the Windows computer,
410
00:31:05,380 --> 00:31:07,380
and then you need to use the Windows computer,
411
00:31:07,380 --> 00:31:09,380
and then you need to use the Windows computer,
412
00:31:09,380 --> 00:31:11,380
and then you need to use the Windows computer,
413
00:31:11,380 --> 00:31:13,380
and then you need to use the Windows computer,
414
00:31:13,380 --> 00:31:15,380
and then you need to use the Windows computer,
415
00:31:15,380 --> 00:31:17,380
and then you need to use the Windows computer,
416
00:31:17,380 --> 00:31:19,380
and then you need to use the Windows computer,
417
00:31:19,380 --> 00:31:23,380
and then it's part of that creation,
418
00:31:23,380 --> 00:31:25,380
canpanas, got Manchester,
419
00:31:25,380 --> 00:31:27,380
came together with Brent Anderson,
420
00:31:27,380 --> 00:31:29,380
and that farmed the whole team.
421
00:31:29,380 --> 00:31:31,380
So I think going back to your question,
422
00:31:31,380 --> 00:31:35,380
what's the best and most impactful feature?
423
00:31:35,380 --> 00:31:39,380
I think that's the decision we made,
424
00:31:39,380 --> 00:31:41,380
in the early stage, to build Windows 365,
425
00:31:41,380 --> 00:31:43,380
and that is inside Microsoft Engine.
426
00:31:43,380 --> 00:31:47,380
Because imagine that, hey, you know Microsoft Engine,
427
00:31:47,380 --> 00:31:51,380
and now it's just a subset of provisioning and cloud PC,
428
00:31:51,380 --> 00:31:53,380
and then all the policies and everything else,
429
00:31:53,380 --> 00:31:55,380
is just merging together.
430
00:31:55,380 --> 00:31:59,380
So I think that's just a very powerful combo.
431
00:31:59,380 --> 00:32:03,380
I think the other parts of what has been very impactful,
432
00:32:03,380 --> 00:32:05,380
is the decision to make,
433
00:32:05,380 --> 00:32:09,380
to make from the Windows 365 app,
434
00:32:09,380 --> 00:32:12,380
and the RDP or the remote desktop app,
435
00:32:12,380 --> 00:32:15,380
you probably know that like old school rat icon,
436
00:32:15,380 --> 00:32:19,380
not on, to merge that together to the Windows app.
437
00:32:19,380 --> 00:32:23,380
And of course, there's a little of debate online,
438
00:32:23,380 --> 00:32:25,380
whether it should be called the Windows app,
439
00:32:25,380 --> 00:32:27,380
because every Windows app that runs on Windows
440
00:32:27,380 --> 00:32:29,380
is called the Windows app,
441
00:32:29,380 --> 00:32:31,380
so it could be very confusing sometimes to use this,
442
00:32:31,380 --> 00:32:35,380
but if you think about the story and the vision of the Windows app,
443
00:32:35,380 --> 00:32:37,380
connecting to Windows and the cloud,
444
00:32:37,380 --> 00:32:39,380
and that's the only Windows app you need in order to do so,
445
00:32:39,380 --> 00:32:41,380
and make complete sense.
446
00:32:41,380 --> 00:32:46,380
So having a unified app now for AVD, Windows 365, RDP,
447
00:32:46,380 --> 00:32:49,380
and all kinds of other features and services,
448
00:32:49,380 --> 00:32:51,380
merging into that coming together,
449
00:32:51,380 --> 00:32:54,380
I think has been another very impactful one,
450
00:32:54,380 --> 00:32:58,380
that has been creating a little of satisfaction amongst users,
451
00:32:58,380 --> 00:33:02,380
to really, yeah, streamline that end-use experience,
452
00:33:02,380 --> 00:33:06,380
because nobody wants to have like separate clients
453
00:33:06,380 --> 00:33:10,380
for separate solutions they use, right?
454
00:33:10,380 --> 00:33:13,380
You don't want to confuse the user,
455
00:33:13,380 --> 00:33:15,380
but hey, you use this app,
456
00:33:15,380 --> 00:33:20,380
not all of a sudden, you need to use this app for these apps.
457
00:33:20,380 --> 00:33:23,380
So it's very, yeah, very confusing to explain that.
458
00:33:23,380 --> 00:33:26,380
So having one app that you can just use,
459
00:33:26,380 --> 00:33:29,380
and then streamline in the app where you go to,
460
00:33:29,380 --> 00:33:32,380
as of course, it might better experience.
461
00:33:32,380 --> 00:33:36,380
Yeah, things are really interesting.
462
00:33:36,380 --> 00:33:40,380
The another big topic where you end,
463
00:33:40,380 --> 00:33:45,380
it's the AI Copilot, AI Agent AI topic.
464
00:33:45,380 --> 00:33:50,380
How different is this,
465
00:33:50,380 --> 00:33:55,380
this is a gender AI from the standard Copilot?
466
00:33:55,380 --> 00:34:03,380
Yeah, I think, well, I think the UI for AI is still an evolution.
467
00:34:03,380 --> 00:34:08,380
I feel I think the UI for AI is no UI in my opinion.
468
00:34:08,380 --> 00:34:12,380
It's more a most minimal UI on the front end
469
00:34:12,380 --> 00:34:17,380
with the maximum outcome as kind of the back end
470
00:34:17,380 --> 00:34:20,380
and the models working on what you want to achieve
471
00:34:20,380 --> 00:34:24,380
and push that back to you and create what you want to do,
472
00:34:24,380 --> 00:34:28,380
what it's code or dark or presentation or something else
473
00:34:28,380 --> 00:34:31,380
or just a markdown file.
474
00:34:31,380 --> 00:34:36,380
But I think the evolution we have been going through is AI
475
00:34:36,380 --> 00:34:42,380
being an assistant on the site, like let's say Copilot or Claude,
476
00:34:42,380 --> 00:34:46,380
just open it up and type from something in or talk with it
477
00:34:46,380 --> 00:34:53,380
and then create something to a multi-agent like orchestrator
478
00:34:53,380 --> 00:34:58,380
that, hey, you want to do this and that simultaneously.
479
00:34:58,380 --> 00:35:03,380
And then build a whole code base or a set of capabilities
480
00:35:03,380 --> 00:35:07,380
or respond to things in social media or your email
481
00:35:07,380 --> 00:35:11,380
or categorize things in your email completely automated.
482
00:35:11,380 --> 00:35:14,380
That part, I think, is happening right now.
483
00:35:14,380 --> 00:35:19,380
And time will tell how fast that is going to be embraced by customers
484
00:35:19,380 --> 00:35:24,380
because you have now Agent 365 and a new capabilities
485
00:35:24,380 --> 00:35:28,380
like Winner 365 for agents that kind of embraces that
486
00:35:28,380 --> 00:35:32,380
and builds kind of a new chapter in this age of AI
487
00:35:32,380 --> 00:35:35,380
of how we do interaction with computers and orchestration
488
00:35:35,380 --> 00:35:41,380
and security and I think that the part that I think is important there
489
00:35:41,380 --> 00:35:44,380
is that we have the traditional software, the GUIs,
490
00:35:44,380 --> 00:35:47,380
the layers of how we interact with systems
491
00:35:47,380 --> 00:35:50,380
that haven't been changing for like 20, 30 years
492
00:35:50,380 --> 00:35:55,380
to changing to completely augmented kind of automation
493
00:35:55,380 --> 00:36:01,380
and that could be MCP or MCP enabled AI first apps
494
00:36:01,380 --> 00:36:05,380
or a computer using Agent or the AI,
495
00:36:05,380 --> 00:36:10,380
the model goes straight into the database like we set, of course, too.
496
00:36:10,380 --> 00:36:13,380
So the GUI gets wrapped out completely
497
00:36:13,380 --> 00:36:18,380
and then the agent is kind of talking directly with the data
498
00:36:18,380 --> 00:36:21,380
and I think that's pretty much right now
499
00:36:21,380 --> 00:36:25,380
when new models like Fable is the neckt big wave
500
00:36:25,380 --> 00:36:28,380
in terms of AI innovation in my opinion
501
00:36:28,380 --> 00:36:31,380
and time will tell how fast that is going.
502
00:36:31,380 --> 00:36:36,380
As the technology is there, the technology can do much more
503
00:36:36,380 --> 00:36:40,380
that currently the business processes of companies are doing.
504
00:36:40,380 --> 00:36:44,380
So I think that the shift there is like,
505
00:36:44,380 --> 00:36:48,380
technology is here, businesses are here
506
00:36:48,380 --> 00:36:50,380
and then of course you have folks like yourself
507
00:36:50,380 --> 00:36:54,380
being early innovators like the frontier innovators
508
00:36:54,380 --> 00:36:57,380
that really go fast but that's still like a handful
509
00:36:57,380 --> 00:37:01,380
or let's say 10, 20% of the big kind of IT industry
510
00:37:01,380 --> 00:37:05,380
arrests is still using co-coded assistance,
511
00:37:05,380 --> 00:37:08,380
kind of UI components in their daily habits
512
00:37:08,380 --> 00:37:12,380
and then moving to agents but still like a human
513
00:37:12,380 --> 00:37:16,380
and the loop kind of agent kind of human and control
514
00:37:16,380 --> 00:37:20,380
agents is doing the work but very much on an individual
515
00:37:20,380 --> 00:37:22,380
one-to-one basis.
516
00:37:22,380 --> 00:37:27,380
The next big wave is doing multi-agent kind of parallel
517
00:37:27,380 --> 00:37:31,380
kind of creation of agents and then autonomous management
518
00:37:31,380 --> 00:37:34,380
and and layers in there are involved.
519
00:37:34,380 --> 00:37:37,380
And I think that's a very exciting space to be in
520
00:37:37,380 --> 00:37:41,380
and that's where I feel that in tune is evolving to as well.
521
00:37:41,380 --> 00:37:45,380
And 365 too with agent 365.
522
00:37:45,380 --> 00:37:50,380
Yeah, I have looked a little bit in your employee Nurdio
523
00:37:50,380 --> 00:37:54,380
and there are some really interesting headlines.
524
00:37:54,380 --> 00:38:01,380
This one is from a PR. Nurdio brings AI to Microsoft Cloud Management
525
00:38:01,380 --> 00:38:06,380
and the other Nurdio Manager for Enterprise name.
526
00:38:06,380 --> 00:38:09,380
This is the, yeah, no, this was the wrong one.
527
00:38:09,380 --> 00:38:12,380
So, I think that's the video enhanced Windows 365
528
00:38:12,380 --> 00:38:14,380
management with AI.
529
00:38:14,380 --> 00:38:19,380
So, what is AI operator agent IT departments look like?
530
00:38:19,380 --> 00:38:23,380
Yeah, well, I think that question,
531
00:38:23,380 --> 00:38:27,380
to ask to that question is kind of variable.
532
00:38:27,380 --> 00:38:32,380
I think from my perspective, what I foresee that's happening
533
00:38:32,380 --> 00:38:36,380
is that traditional UI will still be there for a while
534
00:38:36,380 --> 00:38:38,380
for a couple of years.
535
00:38:38,380 --> 00:38:41,380
And the effect is happening is that part of management
536
00:38:41,380 --> 00:38:44,380
will consolidate to a more unified UI.
537
00:38:44,380 --> 00:38:47,380
So, I met in that you embrace CLAW
538
00:38:47,380 --> 00:38:50,380
or you embrace co-pilot.
539
00:38:50,380 --> 00:38:55,380
Then most likely to co-pilot UI is the one that you use the most
540
00:38:55,380 --> 00:38:57,380
on a daily basis to do your work.
541
00:38:57,380 --> 00:39:02,380
So, what if you want to manage Cloud PCs, physical PCs,
542
00:39:02,380 --> 00:39:07,380
ABD, Android 65, all these things?
543
00:39:07,380 --> 00:39:10,380
Do you then want to kind of step out of the co-pilot UI
544
00:39:10,380 --> 00:39:13,380
you use for 80% of your daily habit, kind of work,
545
00:39:13,380 --> 00:39:15,380
and then go to a traditional one?
546
00:39:15,380 --> 00:39:16,380
I think yes, for you.
547
00:39:16,380 --> 00:39:19,380
Oh, you want to kind of connect that all to get in
548
00:39:19,380 --> 00:39:24,380
that single point of glass view in that AI assistant layer.
549
00:39:24,380 --> 00:39:27,380
And that's what Nurdio, what we are building there,
550
00:39:27,380 --> 00:39:30,380
is really kind of stringlining that MCP enable
551
00:39:30,380 --> 00:39:34,380
the autonomous management layer of booking that in on whatever
552
00:39:34,380 --> 00:39:36,380
front end of UI you want to have,
553
00:39:36,380 --> 00:39:41,380
Cloud and Tropic, OpenAI, co-pilot, it doesn't matter.
554
00:39:41,380 --> 00:39:45,380
But then on the backend, it automates that and streamlines
555
00:39:45,380 --> 00:39:47,380
that for you so the complexity will be gone.
556
00:39:47,380 --> 00:39:52,380
What I feel is imagine you are like an IT manager
557
00:39:52,380 --> 00:39:56,380
and a decision maker and you walk into the IT department
558
00:39:56,380 --> 00:39:58,380
and you are asking normally,
559
00:39:58,380 --> 00:40:01,380
hey, can somebody give me a report of the capability
560
00:40:01,380 --> 00:40:05,380
of the finance department last week because the IT
561
00:40:05,380 --> 00:40:09,380
or the finance manager complained to IT saying
562
00:40:09,380 --> 00:40:12,380
a couple of the users couldn't log on
563
00:40:12,380 --> 00:40:15,380
because there was an error or something occurring.
564
00:40:15,380 --> 00:40:18,380
So, no, there's a very normal thing that happens sometimes
565
00:40:18,380 --> 00:40:21,380
that they want to know what's going on.
566
00:40:21,380 --> 00:40:26,380
Imagine that the IT manager just goes straight to co-pilot
567
00:40:26,380 --> 00:40:31,380
and asks, hey, can you create a report for this department
568
00:40:31,380 --> 00:40:36,380
for me by including these time working hours
569
00:40:36,380 --> 00:40:38,380
and create a PDF for me?
570
00:40:38,380 --> 00:40:40,380
And then it does it completely.
571
00:40:40,380 --> 00:40:43,380
And then what if it can also find a root cause
572
00:40:43,380 --> 00:40:46,380
and then send it straight to the support desk as well
573
00:40:46,380 --> 00:40:49,380
or to the IT department to give them a heads-up
574
00:40:49,380 --> 00:40:54,380
if they couldn't see it or haven't received that message
575
00:40:54,380 --> 00:40:58,380
or the availability concerns or something like that.
576
00:40:58,380 --> 00:41:01,380
So, I think that's the part of where it is
577
00:41:01,380 --> 00:41:03,380
traditional system is evolving to.
578
00:41:03,380 --> 00:41:07,380
So, GUI and the reports and making things visible
579
00:41:07,380 --> 00:41:08,380
is still very important.
580
00:41:08,380 --> 00:41:11,380
But I think the UI will evolve more to,
581
00:41:11,380 --> 00:41:14,380
yeah, personally, UI, did you want to see it at time?
582
00:41:14,380 --> 00:41:17,380
So, you promised, for example, can you create this dashboard
583
00:41:17,380 --> 00:41:19,380
and then it creates a reporting dashboard
584
00:41:19,380 --> 00:41:22,380
where exactly the metrics you're looking for
585
00:41:22,380 --> 00:41:25,380
as you can show and look into based on the exact data
586
00:41:25,380 --> 00:41:27,380
coming from your environment.
587
00:41:27,380 --> 00:41:30,380
And then after that, you can maybe save that for later
588
00:41:30,380 --> 00:41:33,380
or just destroy it and then you recreate it next time
589
00:41:33,380 --> 00:41:34,380
when you need it again.
590
00:41:34,380 --> 00:41:38,380
And that intuitive, kind of adaptive, kind of UI
591
00:41:38,380 --> 00:41:42,380
is something I foresee and it's already happening right now
592
00:41:42,380 --> 00:41:46,380
where a partner is in this ecosystem is evolving to.
593
00:41:46,380 --> 00:41:50,380
And you see it with Agent 265 a little bit happening right now.
594
00:41:50,380 --> 00:41:55,380
But I think the part that is missing there is the adaptive
595
00:41:55,380 --> 00:42:00,380
and legacy and backward compatibility
596
00:42:00,380 --> 00:42:04,380
part of Agent 365 towards other solutions
597
00:42:04,380 --> 00:42:05,380
that customers are using.
598
00:42:05,380 --> 00:42:09,380
I think Agent 365 is amazing for if you're completely
599
00:42:09,380 --> 00:42:14,380
in the native and 365 share point and all these
600
00:42:14,380 --> 00:42:15,380
set of features.
601
00:42:15,380 --> 00:42:18,380
But what if you still have an older server,
602
00:42:18,380 --> 00:42:21,380
an older legacy component running in your own data center,
603
00:42:21,380 --> 00:42:22,380
like older software?
604
00:42:22,380 --> 00:42:26,380
And I think that's the part where I think
605
00:42:26,380 --> 00:42:31,380
the most attention is needed now for IT departments
606
00:42:31,380 --> 00:42:33,380
as well as consultants like MBPs,
607
00:42:33,380 --> 00:42:36,380
like really help there in that journey to AI.
608
00:42:36,380 --> 00:42:40,380
And that's not different from on-prem to cloud.
609
00:42:40,380 --> 00:42:43,380
And like all these other transitions that happened
610
00:42:43,380 --> 00:42:45,380
in the last 20, 30 years,
611
00:42:45,380 --> 00:42:50,380
I guess a previous point in time and there is a new point in time.
612
00:42:50,380 --> 00:42:52,380
How does this kind of intersect?
613
00:42:52,380 --> 00:42:55,380
That's where yeah, partners, consultants,
614
00:42:55,380 --> 00:42:56,380
all come into play.
615
00:42:56,380 --> 00:42:59,380
The part that I haven't seen happening yet
616
00:42:59,380 --> 00:43:07,380
in a very large kill is governance in terms of like training
617
00:43:07,380 --> 00:43:12,380
and help on how can we help a large enterprise
618
00:43:12,380 --> 00:43:15,380
to train to use AI the best way possible.
619
00:43:15,380 --> 00:43:18,380
And we used to have that with Office 365.
620
00:43:18,380 --> 00:43:24,380
I'm 365 and it feels like when we went from exchange
621
00:43:24,380 --> 00:43:28,380
to exchange the line and you probably remember that
622
00:43:28,380 --> 00:43:32,380
time frame beat pause to Office 365.
623
00:43:32,380 --> 00:43:37,380
There were businesses built around explaining
624
00:43:37,380 --> 00:43:39,380
how it works to users.
625
00:43:39,380 --> 00:43:43,380
They came into businesses, they trained like three, four weeks
626
00:43:43,380 --> 00:43:46,380
and then they went out and then it was done.
627
00:43:46,380 --> 00:43:51,380
And I feel that right now the technology is going extremely fast
628
00:43:51,380 --> 00:43:55,380
but the explanation of how you should use it is lacking.
629
00:43:55,380 --> 00:43:58,380
And we know it is we were like tax-abry, we are an IT.
630
00:43:58,380 --> 00:44:01,380
So we are like like learning fast.
631
00:44:01,380 --> 00:44:03,380
And even we cannot keep up sometimes.
632
00:44:03,380 --> 00:44:06,380
So imagine like a normal user in a business
633
00:44:06,380 --> 00:44:10,380
that person's like primary focus isn't IT.
634
00:44:10,380 --> 00:44:12,380
Like they are struggling right now.
635
00:44:12,380 --> 00:44:17,380
So I feel that like what is a very large kind of set of an
636
00:44:17,380 --> 00:44:21,380
opportunity that needs to happen that customers company
637
00:44:21,380 --> 00:44:26,380
need to jump into and community is explaining the A2B,
638
00:44:26,380 --> 00:44:30,380
A2B, C, or let's say chapter one, two and three,
639
00:44:30,380 --> 00:44:35,380
kind of what's in between and how can I modernize towards AI
640
00:44:35,380 --> 00:44:39,380
and how can I create a solution and this is where partners kick in
641
00:44:39,380 --> 00:44:44,380
to overlay it at all and kind of create a migration
642
00:44:44,380 --> 00:44:46,380
that doesn't feel like a migration.
643
00:44:46,380 --> 00:44:50,380
So you can kind of lift and shift between chapter one, two and three.
644
00:44:50,380 --> 00:44:53,380
And maybe from chapter three back to two if it needs to
645
00:44:53,380 --> 00:44:55,380
and then back from two to three.
646
00:44:55,380 --> 00:44:59,380
I think that's a, yeah, future software products
647
00:44:59,380 --> 00:45:03,380
where we are big opportunities in terms of helping customers there.
648
00:45:03,380 --> 00:45:08,380
So will you clean the traditional admin role will change
649
00:45:08,380 --> 00:45:11,380
to a multi agent manager?
650
00:45:11,380 --> 00:45:18,380
I definitely think that admins will get more like digital admins
651
00:45:18,380 --> 00:45:22,380
added to their like let's say roles in responsibility.
652
00:45:22,380 --> 00:45:26,380
So imagine you have an agent that does like security,
653
00:45:26,380 --> 00:45:31,380
policy management for you, but you still in control of what that agent
654
00:45:31,380 --> 00:45:34,380
is doing. I think that's the important part.
655
00:45:34,380 --> 00:45:40,380
Like that you have agents for specific tasks and not like one agent
656
00:45:40,380 --> 00:45:44,380
that I don't know let's say got mode or local administrative
657
00:45:44,380 --> 00:45:49,380
permission to begin destroy everything and also like from a harmful
658
00:45:49,380 --> 00:45:53,380
kind of like hacker kind of intruder perspective.
659
00:45:53,380 --> 00:45:55,380
You don't want to have that happening.
660
00:45:55,380 --> 00:45:59,380
So I think that security and governance layer of how to make sure
661
00:45:59,380 --> 00:46:05,380
that an agent that can do some of that more privilege kind of work
662
00:46:05,380 --> 00:46:09,380
is consolidated to only doing that part that you approve it to.
663
00:46:09,380 --> 00:46:14,380
I think that's the next set of what needs to happen in IT management
664
00:46:14,380 --> 00:46:18,380
with our back permissions that are going beyond on what we have been
665
00:46:18,380 --> 00:46:22,380
seeing in the past where like a, a gentig user.
666
00:46:22,380 --> 00:46:28,380
So the entra, agente user, I user, agente ID kind of approach.
667
00:46:28,380 --> 00:46:32,380
You probably have heard of that too is that you now in the new
668
00:46:32,380 --> 00:46:35,380
agent 365 license a bit as well.
669
00:46:35,380 --> 00:46:40,380
And the agent licensing offering of entra you have now something called an
670
00:46:40,380 --> 00:46:44,380
agente user or digital worker concept where you can have somebody
671
00:46:44,380 --> 00:46:48,380
logging on on behalf of you as the user, but it actually isn't you.
672
00:46:48,380 --> 00:46:54,380
So it's kind of a subset like like a subuser of you as a, as a user.
673
00:46:54,380 --> 00:46:57,380
And then can have similar permissions as you have.
674
00:46:57,380 --> 00:47:02,380
And then you can delegate your task to that user to that agent,
675
00:47:02,380 --> 00:47:03,380
agente user.
676
00:47:03,380 --> 00:47:07,380
And I think that's the next wave of what we will see.
677
00:47:07,380 --> 00:47:12,380
But yeah, the software products and such need to kind of adapt to that.
678
00:47:12,380 --> 00:47:17,380
So for example, the whole concept of logging on into windows as an
679
00:47:17,380 --> 00:47:22,380
agente user is kind of something that is out there now,
680
00:47:22,380 --> 00:47:25,380
but it hasn't really been used by many customers yet.
681
00:47:25,380 --> 00:47:32,380
So it's like, like that part of, hey, all these computer systems,
682
00:47:32,380 --> 00:47:35,380
all these ISFees, all these applications, all these things that have been
683
00:47:35,380 --> 00:47:40,380
built in the last 10, 50 years, all have like support contracts of,
684
00:47:40,380 --> 00:47:43,380
hey, I support this operating system.
685
00:47:43,380 --> 00:47:48,380
I support four cores, 12 gigs of RAM as deep like that stuff,
686
00:47:48,380 --> 00:47:54,380
as all being built around one primary user per one operating system.
687
00:47:54,380 --> 00:48:00,380
And now we are moving out of that model to multiple user or let's say
688
00:48:00,380 --> 00:48:06,380
different agente users to an operating system or different systems that do the same.
689
00:48:06,380 --> 00:48:11,380
But the person itself is an agente kind of user, like a computer using agent
690
00:48:11,380 --> 00:48:14,380
model that comes in and does the work for you.
691
00:48:14,380 --> 00:48:19,380
So I think just going back to your question, I feel the exciting card is that
692
00:48:19,380 --> 00:48:27,380
there's a lot of change and it's faster and it's innovative and I think that's cool.
693
00:48:27,380 --> 00:48:34,380
But the part I think is important there is that faster doesn't always mean better.
694
00:48:34,380 --> 00:48:39,380
And faster could also mean it's more expensive.
695
00:48:39,380 --> 00:48:41,380
So where do the ROI?
696
00:48:41,380 --> 00:48:48,380
Can I take out, like for example you for half a day to work on something else
697
00:48:48,380 --> 00:48:53,380
that you normally do and then do the RRI calculation there and then that
698
00:48:53,380 --> 00:48:55,380
justifies the investment.
699
00:48:55,380 --> 00:49:01,380
I think that part sometimes is harder depending on which workload.
700
00:49:01,380 --> 00:49:08,380
But I think from a management layer, from an in tune, when is 265 security.
701
00:49:08,380 --> 00:49:13,380
I think the exciting part there is that the repetitive tasks, the things you do,
702
00:49:13,380 --> 00:49:20,380
that you know where to click on, but you just are the only person that can do it
703
00:49:20,380 --> 00:49:24,380
because you have these permissions to completely automate that with just a single click
704
00:49:24,380 --> 00:49:29,380
or an automated system, I think is the first set of experiences you will see that AI
705
00:49:29,380 --> 00:49:36,380
will start doing on behalf of users and then eventually move to a more organic
706
00:49:36,380 --> 00:49:43,380
way of doing it. But yet the rights and our back and and and and and MCP enable
707
00:49:43,380 --> 00:49:49,380
configuration flows are currently in development and then our customers will
708
00:49:49,380 --> 00:49:54,380
will define the pace of how fast we can move in that direction.
709
00:49:54,380 --> 00:50:00,380
Yeah, I think a little bit about yeah, keep the mind of loving, but we had a lot of
710
00:50:00,380 --> 00:50:09,380
things we have started with Clippy on Windows. He tried to help us. He's at least that I'm a little bit.
711
00:50:09,380 --> 00:50:15,380
I have one year. Yeah. He was such a cool guy. Then we have Cortana.
712
00:50:15,380 --> 00:50:23,380
It's not really helped. And now also, I don't know, on my case, he is also Windows.
713
00:50:23,380 --> 00:50:31,380
And even those co-pilons also got done this longer here. So can AI travel through Windows
714
00:50:31,380 --> 00:50:38,380
device autonomously or is this something we we have to still wait for?
715
00:50:38,380 --> 00:50:47,380
No, I think AI is completely able to analyze data and find the patterns of when normally
716
00:50:47,380 --> 00:50:55,380
a human would take hours or maybe days to find. So the technology is there. I think the challenge
717
00:50:55,380 --> 00:51:02,380
there is is that is the data currently connected to a system that can automate and
718
00:51:02,380 --> 00:51:08,380
resolve it. And I think that's the difference and that needs to kind of get created and
719
00:51:08,380 --> 00:51:14,380
build out and you already see it like support desks and there's already like an
720
00:51:14,380 --> 00:51:21,380
configuration and agent and mcp SDK and Windows that can go into some of the more
721
00:51:21,380 --> 00:51:28,380
kind of logging and telemetry as well. I think the next wave is here to connect it to
722
00:51:28,380 --> 00:51:35,380
like APIs and correlate from that data and then take action from that data as well.
723
00:51:35,380 --> 00:51:41,380
So I do I do think that troubleshooting is probably one of the best use cases for
724
00:51:41,380 --> 00:51:48,380
AI to jump in and help because yeah it's like analyzing data and then based on that
725
00:51:48,380 --> 00:51:56,380
result what comes out of that it can take an action to resolve it. And normally first
726
00:51:56,380 --> 00:52:01,380
it needs to happen a couple times you learn from that and then you manually create a script
727
00:52:01,380 --> 00:52:07,380
and then have support teams doing that. I think where it's going to is that you're
728
00:52:07,380 --> 00:52:15,380
client where you log into or your desktop your Winnus PC has like an integrated
729
00:52:15,380 --> 00:52:20,380
experience that you can type in and say normally what you normally will say to a
730
00:52:20,380 --> 00:52:27,380
support team. I kind of like use my app it gives me this warning put in a screen
731
00:52:27,380 --> 00:52:33,380
shot of the warning and then AI will resolve it on the back end completely for you.
732
00:52:33,380 --> 00:52:40,380
And the part that currently is not there in Windows is that like m3 65 coal pie doesn't
733
00:52:40,380 --> 00:52:45,380
have that automatically kind of autonomous resolution part India yet. And that's
734
00:52:45,380 --> 00:52:52,380
I think a little bit of a shadow AI shadow IT kind of use cases while because as
735
00:52:52,380 --> 00:52:57,380
a IT professional you don't want to have your users doing troubleshooting without
736
00:52:57,380 --> 00:53:03,380
you giving the consent for them to do so or at least not without you knowing what's
737
00:53:03,380 --> 00:53:07,380
going on or what they probably should then needs to be logging and somewhat of
738
00:53:07,380 --> 00:53:12,380
visibility there and I think that's the part that a little large companies are
739
00:53:12,380 --> 00:53:17,380
figuring out right now and that's where digital employee experience or DAX in
740
00:53:17,380 --> 00:53:23,380
short becomes very important because DAX is kind of kind of the overall employee
741
00:53:23,380 --> 00:53:29,380
experience so imagine AI is doing everything for you then the employee experience is
742
00:53:29,380 --> 00:53:34,380
still important so how do you measure that how do you connect the dots of all
743
00:53:34,380 --> 00:53:41,380
these systems together and still visualize it to see are my users still satisfied
744
00:53:41,380 --> 00:53:47,380
with the experience they have even though everything runs but AI or technology is
745
00:53:47,380 --> 00:53:54,380
great I think the opinion of the user still remains to be important and that's
746
00:53:54,380 --> 00:53:59,380
where you are a digital employee experience that DAX kind of is a new concept
747
00:53:59,380 --> 00:54:05,380
comes in and I feel that that's for like the future another angle of how systems
748
00:54:05,380 --> 00:54:12,380
and traditional concepts are changing in a more experienced focus manner versus
749
00:54:12,380 --> 00:54:18,380
a more hey look at my fancy technology or solution that I use that did this to
750
00:54:18,380 --> 00:54:24,380
more like from the outcome perspective look at the user experience I think that's
751
00:54:24,380 --> 00:54:32,380
that's very important I think when we look a little bit into the
752
00:54:32,380 --> 00:54:38,380
the admin groups and so on we have arguments talk about entra they talk about
753
00:54:38,380 --> 00:54:46,380
peer mute they talk about yeah also AI they talk about defender but there's
754
00:54:46,380 --> 00:54:53,380
it's yeah some tools is a little bit missing I think that into it's still living
755
00:54:53,380 --> 00:55:01,380
well you have to imagine that in tune is is out there for like roughly ten
756
00:55:01,380 --> 00:55:06,380
years now it's sorry it's still the light and then it evolved to the new
757
00:55:06,380 --> 00:55:11,380
in tune portal kind of experience that still today is the portal experience
758
00:55:11,380 --> 00:55:16,380
I just took recall windows in tune as well I know I don't know if you if you're
759
00:55:16,380 --> 00:55:21,380
familiar but windows in tune was the original day the original name in the first
760
00:55:21,380 --> 00:55:28,380
couple I kind of year and since then like since covid in tune grow massively
761
00:55:28,380 --> 00:55:34,380
like hundreds of millions of endpoints are managed by in tune they're still a
762
00:55:34,380 --> 00:55:42,380
big estate on SCCM like large enterprises that still use that and the more a
763
00:55:42,380 --> 00:55:49,380
product grows the more yeah like certain features bigger features are being
764
00:55:49,380 --> 00:55:55,380
done but also how how things are being prioritized are changing as well so
765
00:55:55,380 --> 00:56:00,380
what you see a lot of security copilot and and in tune agents being built
766
00:56:00,380 --> 00:56:07,380
right now recently there were were like some improvements on the speed of
767
00:56:07,380 --> 00:56:12,380
applying policies and apps like with the sidecar management extension
768
00:56:12,380 --> 00:56:16,380
improvements called product fast lane so there's definitely attention for
769
00:56:16,380 --> 00:56:20,380
features that have kind of asked for the for the core set of capabilities of
770
00:56:20,380 --> 00:56:26,380
in tune but I'm hearing it sometimes more at does more attention for AI and less on the
771
00:56:26,380 --> 00:56:31,380
engine actually if you ask for a feature that it's not being shipped and I think
772
00:56:31,380 --> 00:56:39,380
yeah that's well acknowledged and yeah I think the part of partner solutions is
773
00:56:39,380 --> 00:56:45,380
then important that partner solutions help jump again and creating solutions that
774
00:56:45,380 --> 00:56:50,380
can help customers filling into blanks what in tune doesn't do and that's
775
00:56:50,380 --> 00:56:55,380
where like an earlier does a lot of things on all see management backup sequencing
776
00:56:55,380 --> 00:57:00,380
and in tune things like that have or should have been an in tune as well and
777
00:57:00,380 --> 00:57:06,380
that's the beauty of this ecosystem ecosystem as well right like Microsoft
778
00:57:06,380 --> 00:57:12,380
tends to kind of normally do 70% of the core and leave the other 20 to 30%
779
00:57:12,380 --> 00:57:18,380
open for partners to jump in and help customers and build on top of the native
780
00:57:18,380 --> 00:57:27,380
technology stack to you so I think yeah that's something to always be aware of on how
781
00:57:27,380 --> 00:57:31,380
you decide which direction you go but luckily there are more solutions in this
782
00:57:31,380 --> 00:57:38,380
ecosystem as well. Yeah we have I think too powerful tools to write what I don't know
783
00:57:38,380 --> 00:57:44,380
we have PowerShell and Graph API and yeah we have this I think or another
784
00:57:44,380 --> 00:57:54,380
another possible last time it's infrastructure as code that you think in
785
00:57:54,380 --> 00:58:07,380
in tune will also become more a idron in the future. Yeah I do I do think so.
786
00:58:07,380 --> 00:58:15,380
Of course because in tune is so large and has many many many microservices on the
787
00:58:15,380 --> 00:58:22,380
back that are automated and globally distributed those all need to be evolved
788
00:58:22,380 --> 00:58:30,380
to like systems that can talk and automate with like a new front end so imagine
789
00:58:30,380 --> 00:58:39,380
that within Microsoft when a product is being built like m365 works on substrate
790
00:58:39,380 --> 00:58:46,380
and in tune works on graph and in tune is built on other backend services then for
791
00:58:46,380 --> 00:58:53,380
example other services that are on m365 so so I think one needs to happen is
792
00:58:53,380 --> 00:58:59,380
that more of these systems need to kind of consolidate and then an i.e. professional
793
00:58:59,380 --> 00:59:06,380
that currently use in tune as a console could perfectly find in the future maybe
794
00:59:06,380 --> 00:59:12,380
use m365 copilot or an embedded copilot in an operating system or a mobile app
795
00:59:12,380 --> 00:59:18,380
and then do the the management directly from there and I think right now there's
796
00:59:18,380 --> 00:59:23,380
kind of a disconnect between those and I think that needs to be solved and right
797
00:59:23,380 --> 00:59:30,380
now security copilot is the copilot for tune but based on experiences and talking
798
00:59:30,380 --> 00:59:36,380
with customers is that they are not using security copilot the same as m365 copilot
799
00:59:36,380 --> 00:59:41,380
so you have like two copilot there are more copias but let imagine for inch of two copilot
800
00:59:41,380 --> 00:59:46,380
it's one for entry 65 and one for security copilot and they are not kind of blended
801
00:59:46,380 --> 00:59:51,380
together so I feel that's that's where I think the opportunity lays is that
802
00:59:51,380 --> 00:59:56,380
those copias need to blend together and you need to move more to one copilot
803
00:59:56,380 --> 01:00:01,380
experience for like i.e. professionals and all its workers doesn't matter
804
01:00:01,380 --> 01:00:09,380
but based on your role or your agentic kind of workload is evolved the way how it interacts with you
805
01:00:09,380 --> 01:00:17,380
and yeah as you know like you can even use the UI of AI I feel becomes
806
01:00:17,380 --> 01:00:22,380
there is a lot of relevant like it would be as minimalistic as it can be
807
01:00:22,380 --> 01:00:26,380
and outcome what you want to achieve is the most important thing and that could be
808
01:00:26,380 --> 01:00:32,380
by talking with AI or typing with AI but the UI is less relevant
809
01:00:32,380 --> 01:00:37,380
unless you need to show something or present something but then you can create it directly out of it
810
01:00:37,380 --> 01:00:42,380
so I think yeah the systems and how it unifies need to change
811
01:00:42,380 --> 01:00:49,380
but I'm not sure how fast that will happen but recently the leader behind
812
01:00:49,380 --> 01:00:57,380
in tune got got purfew as well as a team so purfew in tune are now two orgs
813
01:00:57,380 --> 01:01:03,380
in one so that might change of course purfew is not the whole security copilot stack
814
01:01:03,380 --> 01:01:08,380
so that's much broader but it's the starting point of maybe more and more
815
01:01:08,380 --> 01:01:13,380
unification that might happen in the future as part of in tune
816
01:01:13,380 --> 01:01:21,380
did you think that the graph i.P.I has a future or will it all become mcp
817
01:01:21,380 --> 01:01:24,380
along that mcp in the future?
818
01:01:24,380 --> 01:01:33,380
I think mcp could hook into graph so you build an mcp that kind of connects
819
01:01:33,380 --> 01:01:38,380
to the backend of in tune over graph and that could work out as well
820
01:01:38,380 --> 01:01:44,380
so I do think there is more future for graph than there is for graphical user interfaces
821
01:01:44,380 --> 01:01:51,380
but the time that evolution is going to be the mandatory way of doing
822
01:01:51,380 --> 01:01:58,380
is really dependent on how fast the habits of users i.T. professionals are going to change
823
01:01:58,380 --> 01:02:01,380
meaning that going back to what I just said earlier
824
01:02:01,380 --> 01:02:06,380
I think there needs to be more focus on how you can do your best job
825
01:02:06,380 --> 01:02:11,380
with i in a different way to explain that and train people on that
826
01:02:11,380 --> 01:02:15,380
and how to do it or how to build it like themselves
827
01:02:15,380 --> 01:02:20,380
like that part is missing right now and I think time will tell how fast
828
01:02:20,380 --> 01:02:27,380
that will evolve in a natural way but if you explain like solution x
829
01:02:27,380 --> 01:02:34,380
as the better solution than solution like y or a where you currently are invested in
830
01:02:34,380 --> 01:02:39,380
there needs to be a story in between how to get from a to b
831
01:02:39,380 --> 01:02:44,380
right and the fund of solution 1 to solution 2 and I feel that's important
832
01:02:44,380 --> 01:02:49,380
to tell that story like explain to customers how they get to ai
833
01:02:49,380 --> 01:02:54,380
and how to evolve their processes as well as their users
834
01:02:54,380 --> 01:02:59,380
on how to interact with systems to ai and it feels sometimes that
835
01:02:59,380 --> 01:03:07,380
we're jumping a couple chapters as I said before and I think that going back to your graph question
836
01:03:07,380 --> 01:03:13,380
graph mcp is becoming more and more important but the reality here is
837
01:03:13,380 --> 01:03:17,380
as well that some customers are still using old school gooey
838
01:03:17,380 --> 01:03:21,380
and they are not even looking at graph they don't they know what it is
839
01:03:21,380 --> 01:03:26,380
but they are not using it and that's a gap that needs to get solved
840
01:03:26,380 --> 01:03:31,380
and community articles podcast will help there but I think ultimately
841
01:03:31,380 --> 01:03:37,380
Microsoft and partners need to jump in there because it's a massive opportunity
842
01:03:37,380 --> 01:03:43,380
to embrace ai and companies and explain to the companies how to use ai
843
01:03:43,380 --> 01:03:46,380
the best in the most efficient way possible
844
01:03:46,380 --> 01:03:52,380
what will you think will end-pipe management look in the future?
845
01:03:52,380 --> 01:03:59,380
I honestly think and it depends a little bit on how much legacy you have in your environment
846
01:03:59,380 --> 01:04:03,380
but imagine you're on entra you're pretty much modern
847
01:04:03,380 --> 01:04:08,380
you're like zero trust enabled you have all the capabilities that currently are
848
01:04:08,380 --> 01:04:13,380
some of these buzzwords you just set right you're like your tax-heavy
849
01:04:13,380 --> 01:04:19,380
you're always on the latest and greatest like you're in the Microsoft ecosystem
850
01:04:19,380 --> 01:04:24,380
I think in like two three years from now we will start to see more agents
851
01:04:24,380 --> 01:04:28,380
popping up that can do the management for you
852
01:04:28,380 --> 01:04:34,380
still with allege and IT expertise needed into an IT department
853
01:04:34,380 --> 01:04:41,380
but less in the sense of like sometimes when you work together in a team
854
01:04:41,380 --> 01:04:46,380
there are people that you instruct to do things or you create manuals
855
01:04:46,380 --> 01:04:49,380
and then they go do these manuals for you
856
01:04:49,380 --> 01:04:54,380
so you kind of instruct them and they are more like the executors
857
01:04:54,380 --> 01:04:58,380
kind of of the things that you orchestrate as an let's say IT manager
858
01:04:58,380 --> 01:05:00,380
or like a director
859
01:05:00,380 --> 01:05:04,380
I think the agents that are coming in the next two to three years for
860
01:05:04,380 --> 01:05:08,380
modern management and in tune and other capable I sent it around the things
861
01:05:08,380 --> 01:05:15,380
there are easily automatable like imagine you create a manual
862
01:05:15,380 --> 01:05:23,380
and you feed it to an agent and that agent goes go do the actions based on that manual for you
863
01:05:23,380 --> 01:05:29,380
and automatically calls over teams but end users and explain that manual
864
01:05:29,380 --> 01:05:33,380
and walks through the steps and then knows based on the model how to interact with it
865
01:05:33,380 --> 01:05:38,380
and then directly to use a base in the knowledge they have and I feel that that's already possible right now
866
01:05:38,380 --> 01:05:41,380
so for example if you call Delta Airlines
867
01:05:41,380 --> 01:05:46,380
Delta of course more US based but if you call Delta Airlines to help desk
868
01:05:46,380 --> 01:05:52,380
you get an AI talking with you and it feels very natural and can anticipate
869
01:05:52,380 --> 01:05:55,380
and go left and right in conversations
870
01:05:55,380 --> 01:06:05,380
versus if you didn't say yes or no you've fallen that blank great area of spot of like computer says no
871
01:06:05,380 --> 01:06:13,380
repeat your thing again right? AI is going to change and is changing that in terms of smart
872
01:06:13,380 --> 01:06:18,380
way of interacting just on a augmented neural network kind of manner
873
01:06:18,380 --> 01:06:23,380
and I think that's the exciting part where troubleshooting, interacting and automating
874
01:06:23,380 --> 01:06:28,380
and autonomous management really benefit from and in the next coming years
875
01:06:28,380 --> 01:06:33,380
I think you'll start seeing the first set of experiences coming to life
876
01:06:33,380 --> 01:06:40,380
and let it as partner, first and then Microsoft later or as kind of a combined kind of offering
877
01:06:40,380 --> 01:06:47,380
time will tell but I think that's the next big wave of endpoint management
878
01:06:47,380 --> 01:06:55,380
autonomous management more automated service, self-service now kind of capabilities
879
01:06:55,380 --> 01:07:01,380
that have been in the past available when customer asks or employee asks for this
880
01:07:01,380 --> 01:07:07,380
and then there's a whole flow happening on the back and then people get kind of approval messages
881
01:07:07,380 --> 01:07:12,380
of hey are you are you giving consent for this? Are you giving consent for that?
882
01:07:12,380 --> 01:07:18,380
I feel that's more and more going to happen but then without the technology involves of building that
883
01:07:18,380 --> 01:07:26,380
yourself it's just a prompt away and then you can say hey configure this policy setting for this group
884
01:07:26,380 --> 01:07:33,380
make sure that the exact setting is being a SAP a kind of pressure test
885
01:07:33,380 --> 01:07:37,380
once or twice first before I'm rolling to the bigger kind of team
886
01:07:37,380 --> 01:07:43,380
and then your T department is being used as well for checking things out
887
01:07:43,380 --> 01:07:48,380
and then eventually it will will get completely automated
888
01:07:48,380 --> 01:07:56,380
but I feel that knowledge of the systems are still important for the next two, three to five years
889
01:07:56,380 --> 01:08:02,380
the question however is how many of these very experienced IT department
890
01:08:02,380 --> 01:08:08,380
kind of IT administrators are needed do you need 10 do you need five do you need three
891
01:08:08,380 --> 01:08:16,380
time will tell but I do feel that as an IT professional being very tax savvy
892
01:08:16,380 --> 01:08:24,380
and knowing how to build things is very important these days to build the agents, build the automation
893
01:08:24,380 --> 01:08:29,380
that all gets you to that next step of autonomous management
894
01:08:29,380 --> 01:08:38,380
I think Microsoft has a fair software part and also Microsoft now has a little bit longer
895
01:08:38,380 --> 01:08:48,380
about the Microsoft sphere so we had a hardware part that you think also we have this become a bigger part
896
01:08:48,380 --> 01:08:58,380
that we get secured the ship, what is the chips that is also growing
897
01:08:58,380 --> 01:09:04,380
and in the future and it will be I don't know I think more simpler to get all
898
01:09:04,380 --> 01:09:10,380
yeah, this is two handle this from the hardware side
899
01:09:10,380 --> 01:09:16,380
I think I think what is important to recognize is that AI runs on GPUs
900
01:09:16,380 --> 01:09:23,380
and GPUs are pretty expensive especially now with the RAM being very scarce to get
901
01:09:23,380 --> 01:09:29,380
it's going like through the roof pretty much and it's harder to get GPUs because all the big
902
01:09:29,380 --> 01:09:33,380
cloud vendors need to GPUs in order to build and skill the system
903
01:09:33,380 --> 01:09:41,380
so what you see there is then the price of AI is going up as well because of that
904
01:09:41,380 --> 01:09:48,380
I'll suddenly like looking back to hey my employee costs X per hour
905
01:09:48,380 --> 01:09:57,380
my tokens cost X per how many tokens per workload per parameter that the model supports
906
01:09:57,380 --> 01:10:07,380
kind of what's the ROI of AI and how do I use AI to replace something versus how is a human still cheaper
907
01:10:07,380 --> 01:10:12,380
so I think the solution there at least one of the solutions there is at AI
908
01:10:12,380 --> 01:10:19,380
so you have all these devices out there that are copied at plus PC they have an MPU in there
909
01:10:19,380 --> 01:10:27,380
and they can run like so many tops of like a S&M model kind of interaction
910
01:10:27,380 --> 01:10:33,380
I feel that in the future in the next let's say year or so more and more devices
911
01:10:33,380 --> 01:10:41,380
like let's say copied at plus PCs from two years ago they will move to copied at plus plus devices
912
01:10:41,380 --> 01:10:52,380
meaning that the MPU becomes so powerful that can run an LLM that you now need a cloud or copied it for to run
913
01:10:52,380 --> 01:11:00,380
and then you spin up there are the cloud servers GPUs and that costs more money because they control the price of that
914
01:11:00,380 --> 01:11:06,380
so what if you buy a device for like a thousand or fifteen hundred bucks or euros
915
01:11:06,380 --> 01:11:13,380
and then it has an MPU in there that can run LLM and then you can just fire it off locally
916
01:11:13,380 --> 01:11:20,380
and I think that's the way to kind of kind of control the cost a little bit as well
917
01:11:20,380 --> 01:11:28,380
and then like slowly the systems like there are new chips being built right now, built for AI
918
01:11:28,380 --> 01:11:36,380
there are like companies like that are currently like like building new chips for AI
919
01:11:36,380 --> 01:11:41,380
and that will take time but once they are out the price of chips will go down
920
01:11:41,380 --> 01:11:44,380
and then the price of software will go down as well
921
01:11:44,380 --> 01:11:49,380
so there might be an interim step there where at AI becomes more and more important
922
01:11:49,380 --> 01:11:57,380
so for example there are examples of where you can build an ecosystem of multiple
923
01:11:57,380 --> 01:12:06,380
MPUs or GPUs that run locally to offload things that you do in the cloud
924
01:12:06,380 --> 01:12:11,380
so for example you have you interact with co-pilot you ask something
925
01:12:11,380 --> 01:12:21,380
and then you do that from a machine that has a local GPU that can offload that from that client to your local device
926
01:12:21,380 --> 01:12:26,380
and then it still feels it runs in the cloud but it actually is rendered on the local device
927
01:12:26,380 --> 01:12:35,380
I feel that something like that is happening right now but the software doesn't kind of embrace that yet
928
01:12:35,380 --> 01:12:41,380
and I feel that that might be a next step where AI becomes a little bit more affordable
929
01:12:41,380 --> 01:12:47,380
if you are a developer and you need to use the model all the time
930
01:12:47,380 --> 01:12:53,380
versus you just use it a couple times a day for creating a dark as a knowledge user
931
01:12:53,380 --> 01:13:00,380
and now we're waiting all for the real working quantum CPUs
932
01:13:00,380 --> 01:13:06,380
Quantum is one but I feel that quantum is two to three years out still
933
01:13:06,380 --> 01:13:11,380
I do feel that might be an interim step that will happen first
934
01:13:11,380 --> 01:13:18,380
that will be allowing you to either run models on a CPU or in a GPU
935
01:13:18,380 --> 01:13:24,380
that is more like a GPU that is in a laptop or like a desktop PC
936
01:13:24,380 --> 01:13:32,380
like you can have a gaming PC and that like GPU in there could be perfectly fine to run models on
937
01:13:32,380 --> 01:13:38,380
so imagine you have a laptop and you have an iGPU in there that does the display rendering
938
01:13:38,380 --> 01:13:41,380
like many laptops iGPUs now
939
01:13:41,380 --> 01:13:45,380
what if you can run some part of a model on an iGPU
940
01:13:45,380 --> 01:13:51,380
and then it offloads part of it from the cloud to avoid that you spend or your tokens go through a roof
941
01:13:51,380 --> 01:13:56,380
I think that are parts of efficiency layers that I feel that are happening
942
01:13:56,380 --> 01:14:01,380
the question however is the model as good on that chip
943
01:14:01,380 --> 01:14:04,380
and as it is on that back end coming from the cloud
944
01:14:04,380 --> 01:14:10,380
and I think that's the technical challenge part that is important to get results
945
01:14:10,380 --> 01:14:16,380
to like when you use this device with this hardware can it offload
946
01:14:16,380 --> 01:14:20,380
because hardware and drivers and all these things are different right
947
01:14:20,380 --> 01:14:23,380
so yeah
948
01:14:23,380 --> 01:14:33,380
but I see the new three cherabyte system elements running on the Nvidia GPUs
949
01:14:33,380 --> 01:14:36,380
it's possible but yeah
950
01:14:36,380 --> 01:14:42,380
yep we have the new RTX I think they are called RTX
951
01:14:42,380 --> 01:14:46,380
the new Surface laptops and the npcpu
952
01:14:46,380 --> 01:14:56,380
and they have a CPU/mpa GPU in there that can run which is that like highly amount of token
953
01:14:56,380 --> 01:15:03,380
and the parameter lm70 models or higher even completely on a local device
954
01:15:03,380 --> 01:15:07,380
and I think that's a big change
955
01:15:07,380 --> 01:15:13,380
and you've seen probably Jensen on the Nvidia keynote
956
01:15:13,380 --> 01:15:17,380
like I think it was four weeks ago of GTX
957
01:15:17,380 --> 01:15:23,380
or no yeah I think it's called GTX that's the Nvidia conference every year in the spring
958
01:15:23,380 --> 01:15:29,380
is that that systems everything is going to change while using this chip
959
01:15:29,380 --> 01:15:33,380
meaning that the operating system becomes more
960
01:15:33,380 --> 01:15:36,380
let's say the whole UI of Windows disappears
961
01:15:36,380 --> 01:15:43,380
and the whole way of how Windows interacts with your files and software remains
962
01:15:43,380 --> 01:15:47,380
so then there still needs to be a core layer of Windows involved
963
01:15:47,380 --> 01:15:51,380
but then on top of that like the front end can change
964
01:15:51,380 --> 01:15:59,380
so meaning that could all of a sudden be that your operating system feels like a more like fluent mobile operating system
965
01:15:59,380 --> 01:16:03,380
that you have on your phone but then on your desktop
966
01:16:03,380 --> 01:16:07,380
and I feel that that's an area that we haven't seen much happening
967
01:16:07,380 --> 01:16:12,380
but Google for example created the aluminium operating system
968
01:16:12,380 --> 01:16:15,380
OS that runs on Android
969
01:16:15,380 --> 01:16:20,380
it feels like a Gromo OS but then completely embedded into Gemini
970
01:16:20,380 --> 01:16:28,380
and I think that's again customers define how fast we move in that direction
971
01:16:28,380 --> 01:16:33,380
but I feel it's an example of a pad they are creating for customers to go after
972
01:16:33,380 --> 01:16:39,380
when AI is the primary UI of what you use on a daily basis
973
01:16:39,380 --> 01:16:46,380
so it's an exciting space to be in especially in endpoint management and user computing
974
01:16:46,380 --> 01:16:55,380
I feel that the way how we manage as well as the end user focused systems, experiences, clients are going to change
975
01:16:55,380 --> 01:17:01,380
and the people that have been doing workplace management in companies
976
01:17:01,380 --> 01:17:05,380
they are still important to facilitate that transition
977
01:17:05,380 --> 01:17:09,380
so be ready for that wave to come
978
01:17:09,380 --> 01:17:11,380
it's coming
979
01:17:11,380 --> 01:17:20,380
yeah I hope that's a little bit yeah, in videos
980
01:17:20,380 --> 01:17:25,380
in videos I get a little bit more competitors, reconpettiters
981
01:17:25,380 --> 01:17:34,380
and it's a Google of a Stenzua and yeah I'm there also something in Intel also but yeah
982
01:17:34,380 --> 01:17:36,380
let's have a look in the future
983
01:17:36,380 --> 01:17:43,380
so it's the price go I hope it's get down especially for me as gamer
984
01:17:43,380 --> 01:17:48,380
yeah so let's jump in the rapid fire rounds
985
01:17:48,380 --> 01:17:53,380
I say a short question and you give a short answer
986
01:17:53,380 --> 01:17:57,380
Windows 365 or Virtual Desktop
987
01:17:57,380 --> 01:18:00,380
Windows 365
988
01:18:00,380 --> 01:18:02,380
Intune or configuration manager
989
01:18:02,380 --> 01:18:03,380
Intune
990
01:18:03,380 --> 01:18:06,380
Copilot or agents
991
01:18:06,380 --> 01:18:10,380
Windows 11 or Windows 365
992
01:18:10,380 --> 01:18:21,380
well that's our one because it uses both so I'll win 365 as I believe the OS will change in the future
993
01:18:21,380 --> 01:18:24,380
PowerShell or Graph API
994
01:18:24,380 --> 01:18:28,380
I'm more a Graph person
995
01:18:28,380 --> 01:18:33,380
Teams or Outlook?
996
01:18:33,380 --> 01:18:37,380
what is the first Outlook or Teams or Teams?
997
01:18:37,380 --> 01:18:40,380
Teams definitely Teams
998
01:18:40,380 --> 01:18:42,380
Pitabol or Vicandre?
999
01:18:42,380 --> 01:18:46,380
I will go for Freakon.l's special
1000
01:18:46,380 --> 01:18:50,380
what is the biggest thing I am if you will bust?
1001
01:18:50,380 --> 01:18:53,380
again what's the biggest AI?
1002
01:18:53,380 --> 01:18:56,380
I'm if you will bust
1003
01:18:56,380 --> 01:19:01,380
the the met you mean like what?
1004
01:19:01,380 --> 01:19:11,380
well I think the biggest net around AI right now is that
1005
01:19:11,380 --> 01:19:19,380
AI can create like full blown production code and for large enterprises from scratch
1006
01:19:19,380 --> 01:19:29,380
there's still human interaction and developer and expertise needed in order to ship quality and code to production
1007
01:19:29,380 --> 01:19:39,380
so I think the met that is out there is that yeah you can you can prompt something today and tomorrow you can ship it to 10 million users
1008
01:19:39,380 --> 01:19:44,380
it can can be better and have a better quality and that's just not true
1009
01:19:44,380 --> 01:19:49,380
it's still in full still deep bugging quality checks and things like that
1010
01:19:49,380 --> 01:19:51,380
it's getting better
1011
01:19:51,380 --> 01:20:01,380
it's like a hundred eighty degree change from a year ago so let's say in a year maybe we can retry and maybe it's actually possible to do it completely
1012
01:20:01,380 --> 01:20:04,380
and to do a new podcast
1013
01:20:04,380 --> 01:20:11,380
what's the one habit that makes you successful?
1014
01:20:11,380 --> 01:20:15,380
probably discipline
1015
01:20:15,380 --> 01:20:28,380
don't give up be disciplined of like sometimes when you have something in your brain or you have an idea or you doing something
1016
01:20:28,380 --> 01:20:36,380
it's it's so easy to get distracted by the daily habits and the daily routine and then all of a sudden a couple weeks later
1017
01:20:36,380 --> 01:20:40,380
you're like I was busy so I didn't spend time on that
1018
01:20:40,380 --> 01:20:54,380
I think my my habit or the things that I think I kind of engraved in my way of working is that everything I say I do
1019
01:20:54,380 --> 01:21:01,380
do what you say and say what you do kind of phrase as well I don't know if that's in German as well
1020
01:21:01,380 --> 01:21:11,380
like what you say you kind of you you kind of yeah kind of sign up yourself too
1021
01:21:11,380 --> 01:21:20,380
and I think I I've been doing that like and I could be a trait or a risk as well right because yeah if you if you do that a lot of times
1022
01:21:20,380 --> 01:21:30,380
and you don't finish in a fast pace and you you you you stack up to a month of work and work and work and then you you end up working a lot
1023
01:21:30,380 --> 01:21:39,380
so that's where I think AI really helps me these days so I have a lot of things that I have in my brain that I now give a spot
1024
01:21:39,380 --> 01:21:50,380
and then the other risk is that the people are the the the the systems and the company and the people around you are not there yet as how fast you can go
1025
01:21:50,380 --> 01:21:58,380
so then you have a lot of things that you work on but then the people to catch up on or or create something from that
1026
01:21:58,380 --> 01:22:07,380
go in a slower pace and that's where I hope agents will start helping me more and more to execute some of that work as well in the future
1027
01:22:07,380 --> 01:22:12,380
yeah and when you get all the resources and money you need what future will you develop
1028
01:22:12,380 --> 01:22:32,380
I think honestly the things that I I would be most proud of if if we can pull this off in a good way in the next year or so is a revamp of let's say
1029
01:22:32,380 --> 01:22:46,380
end point management but then completely embraced and infused with AI but then still in the most secure way possible and adaptive way possible where you want to be meaning that I don't you want to go a VD
1030
01:22:46,380 --> 01:22:52,380
you want to go win this 365 you want to go engine you want to go as CCM you want to go
1031
01:22:52,380 --> 01:23:06,380
other solution like doesn't matter like there's a machine a system a machine created that can connect the dots altogether I think that's if we can pull that off in a seamless way
1032
01:23:06,380 --> 01:23:17,380
like I sometimes use the analogy of becoming the Switzerland of multi cloud and AI and and the people in Europe know what that means
1033
01:23:17,380 --> 01:23:29,380
like being kind of universal on the decisions especially right now in the you know on the world and the difference in terms of how people make their investments on cloud
1034
01:23:29,380 --> 01:23:42,380
whether the cloud providers in Europe or in in the US there's like more focused around data sovereignty imagine there's a solution that can connect the dots that you can make the decision
1035
01:23:42,380 --> 01:23:52,380
based on where you want to be at a time but if that changes in six months that can also change back without the complexity involved I think that's
1036
01:23:52,380 --> 01:24:02,380
that's where where I will be investing my money and my team and my resources if I will build something awesome thank you yeah so then my closing question is if
1037
01:24:02,380 --> 01:24:07,380
let's not remember only one thing from today what show the fee
1038
01:24:07,380 --> 01:24:21,380
so the question is if there is one thing from today you will want to share again what would that be is that what you said yeah what is the one thing they they should know
1039
01:24:21,380 --> 01:24:35,380
okay that is get your hands dirty with AI I think the most important thing around AI is start to do and in the standard by doing
1040
01:24:35,380 --> 01:24:52,380
like the brain and this is just some type or like let's say based on research and based on synthetic or synthetic scientific I'm going to say research has been proven out that the brain works better
1041
01:24:52,380 --> 01:25:03,380
if you actually do the things versus only doing the theory the reading the yeah the stuff like I don't know if you're
1042
01:25:03,380 --> 01:25:16,380
if you if you let's say read a manual without actually doing it you forget it like after a week maybe you don't even know what step to look but if you
1043
01:25:16,380 --> 01:25:26,380
performed a manual you probably still know 50 60 70% of the manual a couple months later because you performed the steps you kind of created that
1044
01:25:26,380 --> 01:25:41,380
that memory in your brain that wiring of the things you did to the theory so my advice for everyone would be to get your hand dirty with AI try it out see if it works for you and if not try something different
1045
01:25:41,380 --> 01:25:52,380
and that's the way to get there and the systems have become easier and easier to use so you had a way to ramp up like for example cloth and
1046
01:25:52,380 --> 01:26:06,380
and traffic has been amazing how fast and easy as being evolved to you install it you can create skill sets and agents that are kind of directly attached to you what you can do as
1047
01:26:06,380 --> 01:26:18,380
well as as a company you have an enterprise subscription and then you can pre train the model for everyone in your business meaning that if you miracle will say a
1048
01:26:18,380 --> 01:26:33,380
little bit of the kind of data is key no deck for me based on I don't know when is 365 and an endpoint management that automatically pulls like data from what the company has trained to
1049
01:26:33,380 --> 01:26:45,380
model on as well as the data sources that it's connected to as well as the template and the branding you should use as part of creation of the deck like imagine that normally you have to go to the
1050
01:26:45,380 --> 01:26:59,380
marketing team pull the latest deck review it all these things to just be a prop and then it's there like just examples of how fast things evolve like I know in three months from now things
1051
01:26:59,380 --> 01:27:12,380
could be even more streamlined so so you try it out don't be that last person in line that in choose to pick up AI because yeah of course you don't
1052
01:27:12,380 --> 01:27:23,380
want to be the last want to be efficient but you also want to be yeah prepared for what's next and the moment that regardless of what you do what are
1053
01:27:23,380 --> 01:27:31,380
you're an IT whether you're doing something else but use a eyes part of that day job that will be a moment in time that there will be
1054
01:27:31,380 --> 01:27:46,380
a change happening whether it's new systems being used new opportunities you get a different role things like that if you prepare yourself always for the future you will be always
1055
01:27:46,380 --> 01:27:58,380
one step ahead for the others are in front of the others that don't do that or take another or pro cautious last proactive approach there so I think career wise it's
1056
01:27:58,380 --> 01:28:02,380
a great way to invest in the ideas days as well.
1057
01:28:02,380 --> 01:28:12,380
Yeah so Christian thank you so much for joining me on the m65 and podcast yeah I'm sorry for the technical issues I had today it's
1058
01:28:12,380 --> 01:28:24,380
well if you are frustrating you know the story behind Azure mental test of a Windows 365 but even more exciting was your vision for what's next yeah we explored how
1059
01:28:24,380 --> 01:28:33,380
AI is set for really find it management why autonomous systems will chance form an out operation and how Microsoft build the next generation of
1060
01:28:33,380 --> 01:28:44,380
Intellect Gen work play step boundaries yeah so thank you for for for staying here and if you join this episode be sure to subscribe to
1061
01:28:44,380 --> 01:28:54,380
podcast and you find all links from Christians blocks sessions books and recommendation sources in the show notes so thank you for
1062
01:28:54,380 --> 01:29:01,380
this and thank you for staying here with me Christian thanks man girl I am alright bye bye
1063
01:29:01,380 --> 01:29:03,180
(dog barks)
Founder of m365.fm, m365.show and m365con.net
Mirko Peters is a Microsoft 365 expert, content creator, and founder of m365.fm, a platform dedicated to sharing practical insights on modern workplace technologies. His work focuses on Microsoft 365 governance, security, collaboration, and real-world implementation strategies.
Through his podcast and written content, Mirko provides hands-on guidance for IT professionals, architects, and business leaders navigating the complexities of Microsoft 365. He is known for translating complex topics into clear, actionable advice, often highlighting common mistakes and overlooked risks in real-world environments.
With a strong emphasis on community contribution and knowledge sharing, Mirko is actively building a platform that connects experts, shares experiences, and helps organizations get the most out of their Microsoft 365 investments.
Senior Director, Cloud & AI, Author & Inventor
Through the FLogix acquisition into Microsoft, Christian played a key role in shaping Azure Virtual Desktop and Windows 365 from early concepts into today's multi--billion--dollar cloud virtualization platforms. I contributed to foundational innovations including Windows
365 Boot, Switch, the Windows App, and next--gen Windows Al experiences and an author of 5 books. He's also an inventor on eight patents across cloud, Al, and virtualization technologies that continue to influence modern Cloud PC and endpoint innovation.