Enterprise AI Experiences in Power Apps with Sara Lagerquist [MVP]
In this episode of M365.FM, Mirko Peters talks with Microsoft MVP Sara Lagerquist about how Power Apps is evolving beyond traditional forms and screens into AI-powered, enterprise-ready business experiences. Instead of treating AI as a feature that is simply added to an app, Sara explains how organizations should design applications where AI becomes a natural part of the user experience, helping employees make better decisions, automate repetitive work, and interact with business data more intelligently.
The conversation explores Microsoft's latest innovations, including Generative Pages, Copilot capabilities, Dataverse integration, and modern Power Apps architecture. Sara shares practical guidance on balancing low-code productivity with enterprise requirements such as governance, security, scalability, and maintainability. Rather than focusing on building visually impressive demos, she emphasizes creating business applications that deliver measurable value and remain manageable as organizations grow.
Listeners will also learn how AI is changing the role of makers and developers, why strong data models remain the foundation of successful applications, and how collaboration between citizen developers and professional developers creates the best long-term outcomes. The discussion highlights real-world design patterns, user adoption strategies, and the importance of building solutions that people actually enjoy using.
Whether you're a Power Platform maker, solution architect, IT leader, or Microsoft consultant, this episode offers valuable insights into building modern, AI-driven Power Apps that move beyond simple automation and become intelligent business applications ready for enterprise adoption.
Quick answer: enterprise AI experiences in Power Apps is explained in this M365 FM episode through its purpose, practical use, and the architecture, security, governance, or operational decisions teams should understand before adopting it in a Microsoft environment.
AI plays a crucial role in transforming enterprise applications. It enables businesses to enhance efficiency, make informed decisions, and improve customer interactions. Microsoft Power Apps stands at the forefront of this digital transformation, particularly in building enterprise-ready AI experiences. Organizations leveraging Power Apps have reported impressive results, including a 224% ROI and significant reductions in tech stack expenses. This platform empowers you to build scalable and maintainable AI experiences that align with your unique organizational needs. Embracing these capabilities can pave the way for innovative solutions in your business.
Key Takeaways
- Leverage Microsoft Power Apps to build scalable AI solutions that meet your business needs.
- Define clear business objectives to guide your AI integration and ensure alignment with organizational goals.
- Utilize AI Builder to automate tasks and enhance workflows, improving efficiency and decision-making.
- Incorporate user feedback to continuously improve your AI applications and enhance user satisfaction.
- Focus on security and compliance by adhering to regulations like GDPR and HIPAA to protect user data.
- Engage users with effective training strategies to foster adoption and confidence in using AI-powered applications.
- Implement best practices for continuous improvement to keep your AI solutions effective and relevant.
- Explore the capabilities of Power Apps to stay competitive and enhance operational efficiency in your organization.
Power Apps and AI Capabilities
Key Features of Microsoft Power Apps
Microsoft Power Apps stands out as a transformative platform in enterprise software, offering a suite of features that integrate seamlessly with AI. You can quickly turn ideas into working solutions with AI-assisted suggestions. This allows you to use natural language to define goals and workflows, enabling AI to create structured blueprints. Here are some core features:
- Design secure data models with AI assistance for optimal structure.
- Get recommendations on combining components to solve business challenges.
- Build apps using drag and drop, generative AI, or custom code.
- Accelerate UI design and data modeling with AI assistance.
- Empower users to describe needs in natural language for custom app experiences.
- Integrate Microsoft 365 Copilot for contextual assistance in apps.
- Monitor AI agent activity for better control and oversight.
- Use AI to autofill forms and reduce repetitive tasks.
- Ask apps questions in natural language for quick insights.
These features make Microsoft Power Apps a powerful tool for digital transformation, enabling you to create scalable and maintainable applications that align with your organizational needs.
AI Builder and Automation
Enhancing Workflows with AI
AI Builder in Microsoft Power Apps provides a low-code AI tool that allows you to apply machine learning to automate tasks. It integrates with Power Automate and Power Apps, enhancing applications with smart functionality. Here’s how AI Builder can enhance your workflows:
- Prebuilt models like object detection, text recognition, and sentiment analysis automate processes.
- Custom models allow you to embed machine learning directly into your applications.
- Predictive models help anticipate customer needs and predict business outcomes.
- Natural language processing enables apps to process text and respond to user input intuitively.
- Automation of business processes improves efficiency by automating routine tasks within apps.
Intelligent Decision-Making
The integration of Copilot enhances automation capabilities in Power Apps by providing intelligent suggestions in real-time. Copilot refines interfaces and connects data sources, improving the overall user experience. It acts as a workforce multiplier by automating mundane repetitive tasks. You can interact with Copilot using natural language, making it accessible for users of all skill levels. This integration allows you to build applications faster using natural language prompts, reducing dependency on traditional development efforts.
Building Enterprise-Ready AI Experiences

Defining Business Needs
To build enterprise-ready AI experiences, you must first define your business needs clearly. This process involves several key steps:
- Define business objectives: Identify the outcomes you want to achieve, such as improving productivity or enhancing customer service.
- Assess data readiness: Evaluate the quality, accessibility, and security of your data while establishing governance policies.
- Identify high-value use cases: Focus on targeted use cases that offer measurable impact and manageable complexity.
- Establish governance and security: Implement policies that address security, compliance, and responsible AI usage.
- Build a cross-functional team: Involve stakeholders from various functions to align AI capabilities with organizational priorities.
- Select technologies and platforms: Choose technologies that align with your goals and existing infrastructure.
- Start pilot projects: Validate assumptions and gather feedback through pilot programs.
- Integrate AI into workflows: Embed AI into everyday business processes for maximum impact.
- Support adoption and training: Provide training and establish guidelines to support employee adoption.
- Monitor and improve: Continuously oversee and update systems to maintain effectiveness.
By following these steps, you can ensure that your AI solutions align with your organizational goals and drive meaningful results.
Designing User Interfaces
User interface design plays a crucial role in the success of your AI applications. A well-designed interface enhances user experiences and promotes adoption. Here are some important considerations:
Utilizing Generative Pages
Generative Pages in Power Apps allow you to create user interfaces quickly and efficiently. This feature leverages AI to assist in the design process, enabling you to generate complex interfaces in a fraction of the time. By using generative pages, you can:
- Accelerate development: Save time by automating the creation of user interfaces.
- Enhance customization: Tailor the interface to meet specific business needs and user preferences.
- Improve accessibility: Ensure that your applications are usable by a diverse range of users.
Customization for User Experience
Customization is key to creating engaging user experiences. Here are some design principles to consider:
| Design Principle | Description |
|---|---|
| Behavioral Personalization at the Interface Level | AI adapts the interface based on individual user profiles, enhancing user experience through structural changes. |
| Predictive Navigation and Smart Defaults | AI anticipates user actions, streamlining navigation and reducing cognitive load with smart defaults. |
| Continuous Multivariate Testing | AI automates testing processes, allowing for rapid iteration and optimization of design variants without human bottlenecks. |
| Automated Accessibility Enforcement | AI ensures compliance with accessibility standards by continuously scanning and correcting interface issues. |
| Emotion and Sentiment-Aware Interfaces | AI detects user frustration and responds proactively, enhancing user support and satisfaction. |
| Context-Aware Responsive Design | AI adapts the interface based on contextual factors, providing tailored experiences for different user scenarios. |
By implementing these principles, you can create user interfaces that not only meet functional requirements but also enhance user satisfaction and adoption.
Real-World Applications of AI in Power Apps

Case Study 1: Healthcare Provider Transformation
A leading healthcare provider faced challenges in managing patient records and scheduling appointments. They struggled with inefficiencies that slowed down operations. To address these issues, they developed a custom Power Apps application that digitized patient records. This solution automated appointment scheduling with a user-friendly drag-and-drop calendar interface.
| Key Challenges | Solutions |
|---|---|
| Inefficiencies in patient record management | Developed a custom Power Apps application that digitized patient records. |
| Manual appointment scheduling | Automated scheduling with a drag-and-drop calendar interface and appointment reminders. |
| Lack of centralized data | Integrated data view for doctors linking Power Apps with existing backend systems. |
| Compliance and security risks | Implemented Microsoft Dataverse for secure data storage and ensured HIPAA compliance. |
The results were impressive. Users saved 250 hours per year for high-impact use cases and 10 hours for medium-impact cases. This translated to $31.3 million in value over three years. Additionally, professional developers reduced app development time by 50%, leading to $14.7 million in savings over the same period. This case illustrates how AI in Power Apps can streamline operations and enhance productivity in enterprise applications.
Case Study 2: Retail Inventory Management
In the retail sector, a major company sought to improve its inventory management system. They integrated AI tools within Power Apps to monitor stock levels in real-time. This integration allowed them to trigger actions instantly, enhancing responsiveness to market changes.
The company automated complex workflows, reducing manual intervention in inventory tracking. This automation sped up processes like order fulfillment and minimized errors. The AI-powered assistants enabled non-technical users to create apps and automate tasks using natural language. This democratization of technology empowered employees across the organization.
Furthermore, tools like Process Advisor identified inefficiencies in their operations. They provided recommendations for continuous process optimization, ensuring ongoing enhancement of business operations. As a result, the retail company experienced a significant boost in efficiency and accuracy, leading to improved customer satisfaction and increased sales.
These case studies demonstrate the transformative impact of AI in Power Apps. Organizations can leverage AI to enhance workflows, improve decision-making, and optimize business processes effectively.
Best Practices for AI-Powered Business Applications
Ensuring Security and Compliance
When developing AI-powered business applications, security and compliance are paramount. You must adhere to various regulations to protect user data and maintain trust. Here are some essential standards to consider:
- Compliance with GDPR (General Data Protection Regulation)
- Adherence to HIPAA (Health Insurance Portability and Accountability Act)
- Following CCPA (California Consumer Privacy Act)
- Implementing data protection practices such as encryption and role-based access
- Conducting regular compliance audits to ensure adherence to standards
You should also focus on ethical AI practices. This includes ensuring fairness in AI treatment of users, accountability for AI performance, and transparency in AI operations. By establishing a robust governance framework, you can align your AI solutions with organizational policies and regulatory expectations.
User Training and Adoption Strategies
User adoption is crucial for the success of AI-powered business applications. Engaging users through effective training can significantly enhance their confidence and proficiency. Here are some strategies to consider:
- Use Power BI and other Power Platform tools to track training outcomes and measure employee progress.
- Adjust training content based on measured outcomes to enhance effectiveness.
- Teach users to identify appropriate scenarios for automation or app building to improve usability.
- Implement blended training formats, combining self-paced learning with instructor-led workshops for flexibility.
- Leverage AI to provide personalized training experiences and recommend learning paths.
By focusing on user training, you can foster a culture of adoption. Engaging users through quick wins and intuitive interfaces can lead to improved operational efficiency and better decision-making.
Continuous Improvement Practices
Maintaining AI-powered applications requires ongoing evaluation and enhancement. Here are some best practices for continuous improvement:
| Continuous Improvement Practice | Description |
|---|---|
| User Feedback | Regularly gather feedback from users to identify areas for improvement and iterate on the app's design and functionality. |
| ALM Best Practices | Consistently utilize solutions and distinct environments (development, test, production) to manage app versions effectively. |
| Documentation | Maintain comprehensive documentation of the app's logic, data sources, UI design decisions, and any custom connectors or integrations. |
| Model Retraining | Periodically retrain your Azure ML models with fresh data to ensure their predictions remain accurate and relevant. |
| Security Reviews | Regularly review security roles, data access permissions, and Data Loss Prevention (DLP) policies to ensure ongoing compliance and data protection. |
By implementing these practices, you can ensure that your AI-powered business applications remain effective and aligned with evolving business processes.
Benefits and Challenges of AI Integration
Advantages of AI in Enterprise Applications
Integrating AI into your enterprise applications can significantly enhance efficiency and decision-making. Here are some key advantages:
| Advantage | Description |
|---|---|
| Accelerated Application Development | Power Apps reduces development time by 50%, allowing teams to build applications faster without increasing headcount. |
| Intelligent Automation | Power Automate automates repetitive tasks, leading to a 248% ROI by freeing up knowledge worker time. |
| Improved Decision-Making | AI Builder enables real-time AI-informed decisions by integrating machine learning models directly into applications. |
AI enhances your ability to analyze large volumes of data. It identifies patterns that are hard to detect manually. Predictive analytics helps you forecast trends and evaluate potential outcomes. By automating repetitive tasks, AI allows you to focus on higher-value work. It streamlines workflows, reduces bottlenecks, and improves process consistency. This leads to cost reductions and more effective teamwork. Additionally, AI supports demand forecasting, inventory optimization, and logistics planning. AI systems analyze data from various sources to anticipate disruptions and respond to changing conditions. Tools like chatbots enhance customer service, contributing to overall operational efficiency.
Challenges in Implementation
While the benefits are substantial, you may face several challenges when implementing AI in your enterprise applications.
Technical Limitations
Technical limitations can hinder the scalability of your AI solutions. Here are some common issues:
- Data Quality Issues: Poor data quality can lead to ineffective AI models and erode trust among stakeholders.
- Accessibility Issues: Integration of AI solutions with existing systems can be challenging, impacting scalability.
- Compatibility: Addressing compatibility with legacy systems and performance expectations is essential for scalable AI solutions.
Change Management
Organizational resistance to change can also pose a significant challenge. Employees may fear job displacement or feel uncertain about AI's impact on their roles. To overcome this resistance, consider these strategies:
- Address Job Security: Communicate openly about how AI augments roles and provide clear reskilling pathways.
- Cultivate Experimentation: Create safe 'sandbox' environments for hands-on learning without fear of failure.
- Leverage Peer Advocates: Empower internal influencers to champion the change and build grassroots momentum.
- Ensure Tech Reliability: Be transparent about how AI works, maintain data quality, and keep humans in the loop.
- Establish Ethical Policies: Develop and share clear guidelines on data privacy, fairness, and responsible AI use.
- Listen and Adapt: Actively seek feedback and adjust the process based on legitimate employee concerns.
By addressing these challenges, you can create a smoother transition to AI integration in your enterprise applications.
Building enterprise-ready AI experiences in Power Apps empowers you to transform your business processes. Focus on governance, citizen development, and redesigning workflows to leverage AI effectively. Here are some actionable steps to get started:
- Define your business requirements to identify the specific problem to solve.
- Use Copilot in Power Apps to create your application by entering a prompt for the desired features.
- Connect to relevant data sources like Microsoft Dataverse or SharePoint.
- Add AI features such as document processing and text generation to enhance functionality.
- Automate business processes using Power Automate to streamline workflows.
- Integrate with Microsoft 365 services to improve accessibility and user adoption.
- Publish the application for organizational use across various platforms.
By embracing these strategies, you can stay competitive in the evolving AI landscape and enhance operational efficiency. Explore the capabilities of Microsoft Power Apps today and take the first step towards integrating AI into your business applications.
FAQ
What is Microsoft Power Apps?
Microsoft Power Apps is a low-code platform that allows you to build custom applications quickly. It integrates AI capabilities, enabling you to create scalable and maintainable solutions tailored to your business needs.
How does AI Builder enhance Power Apps?
AI Builder provides low-code tools to integrate AI into your applications. You can automate tasks, analyze data, and create predictive models, enhancing workflows and decision-making processes.
Can I use Power Apps without coding experience?
Yes! Power Apps is designed for users of all skill levels. You can leverage its drag-and-drop interface and AI-assisted features to build applications without extensive coding knowledge.
What types of applications can I create with Power Apps?
You can create various applications, including data entry forms, dashboards, and workflow automation tools. Power Apps supports both canvas and model-driven apps, catering to different business needs.
How does Power Automate work with Power Apps?
Power Automate integrates seamlessly with Power Apps to automate workflows. You can create automated processes that trigger actions based on specific events, improving efficiency and reducing manual tasks.
What are Generative Pages in Power Apps?
Generative Pages are a feature that allows you to quickly create user interfaces using AI assistance. This feature accelerates development and enhances customization, making it easier to meet specific business requirements.
How can I ensure security in my Power Apps applications?
To ensure security, implement governance policies, use role-based access controls, and comply with regulations like GDPR and HIPAA. Regular audits and monitoring also help maintain data protection.
What support resources are available for Power Apps users?
Microsoft offers extensive documentation, tutorials, and community forums for Power Apps users. You can also access training resources and webinars to enhance your skills and knowledge.
π§ Listen to this episode
Want a practical explanation of Enterprise AI Experiences in Power Apps? This episode breaks down the topic in clear language and shows why it matters for Microsoft 365, Azure, Power Platform, security, AI, and modern work.
Listen to this episode if you want to:
- Understand the key concepts behind Enterprise AI Experiences in Power Apps
- See how it fits into the wider Microsoft technology ecosystem
- Learn where it can create practical value for your organization
You may also enjoy these related M365 FM episodes:
- Power Automate vs Azure Logic Apps for Enterprise Integration
- Build Enterprise Automation with Azure Logic Apps
- Microsoft Defender for Cloud Apps - Simply Explained
- Power Apps - Simply Explained
- From Data to Intelligent Agents: Building Trusted Enterprise AI with Microsoft AI Foundry with Shubhangi Goyal [MVP]
Discover more practical Microsoft conversations on M365 FM.
Last reviewed: July 2026.
Who Should Listen
This episode is for Microsoft practitioners, architects, developers, security professionals, and IT leaders evaluating the topic in a real-world environment.
π§ You Should Also Listen To
- AI Agents β A relevant next step that adds practical context to this topic.
- Model Context Protocol β A relevant next step that adds practical context to this topic.
- Microsoft AI Agent Harness β A relevant next step that adds practical context to this topic.
π Want to be part of m365.fm?
Then stop just listening… and start showing up.
π Connect with me on LinkedIn and let’s make something happen:
- ποΈ Be a podcast guest and share your story
- π§ Host your own episode (yes, seriously)
- π‘ Pitch topics the community actually wants to hear
- π Build your personal brand in the Microsoft 365 space
This isn’t just a podcast — it’s a platform for people who take action.
π₯ Most people wait. The best ones don’t.
π Connect with me on LinkedIn and send me a message:
"I want in"
Let’s build something awesome π
1
00:00:00,000 --> 00:00:04,000
>> Yeah, welcome back to the MC65 podcast.
2
00:00:04,000 --> 00:00:06,040
Today, we are driving into one of
3
00:00:06,040 --> 00:00:08,220
the hottest topic in the Microsoft ecosystem,
4
00:00:08,220 --> 00:00:12,920
how AI fundamentally changed the way we build business applications,
5
00:00:12,920 --> 00:00:17,280
while everyone is talking about co-pilot and generative AI.
6
00:00:17,280 --> 00:00:20,920
Very few people are discussing what happens after the demos.
7
00:00:20,920 --> 00:00:25,240
How do we build a iPod business application that are main and
8
00:00:25,240 --> 00:00:29,580
table, governance, sales, and actually resolve real business problems?
9
00:00:29,580 --> 00:00:32,880
Our guest today is someone who works exactly on this space.
10
00:00:32,880 --> 00:00:36,080
Sarah Lagerkwist, she is a Microsoft MVP,
11
00:00:36,080 --> 00:00:38,560
Applications Architect, International Speaker and
12
00:00:38,560 --> 00:00:41,660
Found of the Power Platform Community Sweden.
13
00:00:41,660 --> 00:00:44,880
She specialized in Power Apps, Model Driven Apps,
14
00:00:44,880 --> 00:00:50,360
Custom Pages, PCFs, and extended generative pages
15
00:00:50,360 --> 00:00:52,760
beyond the out-of-the-box experience to create
16
00:00:52,760 --> 00:00:54,660
experience-rate business solutions.
17
00:00:54,660 --> 00:00:57,520
Today, we are all exploring AI,
18
00:00:57,520 --> 00:01:00,460
Architect, governance, Power Platform, and best practice,
19
00:01:00,460 --> 00:01:04,140
Model Driven Apps, the future of enterprise, and so on,
20
00:01:04,140 --> 00:01:05,760
and much, much more.
21
00:01:05,760 --> 00:01:08,320
Welcome Sarah to the show.
22
00:01:08,320 --> 00:01:10,520
>> Thank you. Wow, there was a long entry
23
00:01:10,520 --> 00:01:13,920
within a lot of stuff that I do, but yes, thank you.
24
00:01:13,920 --> 00:01:19,440
>> Yeah, can you tell us a little bit about yourself and
25
00:01:19,440 --> 00:01:22,400
your journey into the Microsoft ecosystem?
26
00:01:22,400 --> 00:01:26,800
>> It's laggy.
27
00:01:26,800 --> 00:01:29,160
>> Absolutely, absolutely.
28
00:01:29,160 --> 00:01:33,520
So I don't have an IT background or
29
00:01:33,520 --> 00:01:37,960
having really studied after the basic school that you have to go through.
30
00:01:37,960 --> 00:01:40,640
I was one of those indecisive persons here and
31
00:01:40,640 --> 00:01:43,680
every decided where, what, a way to go.
32
00:01:43,680 --> 00:01:53,440
So I ended up at a company as an administrator that got CRM 2011 implemented.
33
00:01:53,440 --> 00:01:54,440
>> Okay.
34
00:01:54,440 --> 00:01:59,440
>> That I was just like, huh, this system I kind of get.
35
00:01:59,440 --> 00:02:01,560
This is like, it came natural to me.
36
00:02:01,560 --> 00:02:02,360
I understood it.
37
00:02:02,360 --> 00:02:03,200
It was logical.
38
00:02:03,200 --> 00:02:04,400
It made sense.
39
00:02:04,400 --> 00:02:08,280
I started to do my own workflows and
40
00:02:08,280 --> 00:02:10,440
I found configurations settings and
41
00:02:10,440 --> 00:02:15,120
productions that I shouldn't have found that I did.
42
00:02:15,120 --> 00:02:21,080
And then I looked into like going to school and study something in that direction.
43
00:02:21,080 --> 00:02:25,240
But then they consultancy firm who implemented this said like,
44
00:02:25,240 --> 00:02:27,400
now I just come and work for us instead.
45
00:02:27,400 --> 00:02:32,080
And that's how it's been now for 12 years.
46
00:02:32,080 --> 00:02:33,800
Different firm.
47
00:02:33,800 --> 00:02:36,280
I made a different company now, but yeah,
48
00:02:36,280 --> 00:02:42,080
a consultant for 12 years now pretty much working with a power platform mainly.
49
00:02:42,080 --> 00:02:50,360
>> And why is power platform became your primary focus?
50
00:02:50,360 --> 00:03:01,080
>> Because I think it opens up for supporting more processes than just what the dynamics apps do.
51
00:03:01,080 --> 00:03:07,840
The dynamics apps already have a thought of how the processes should work and
52
00:03:07,840 --> 00:03:09,920
what workflows you should get.
53
00:03:09,920 --> 00:03:15,160
I like the challenge of someone saying, this is how our business works.
54
00:03:15,160 --> 00:03:19,280
And I can help design, I can redesign the process.
55
00:03:19,280 --> 00:03:23,720
I can actually adapt an app or an agent or
56
00:03:23,720 --> 00:03:27,720
something completely based on that.
57
00:03:27,720 --> 00:03:31,640
The data can be different, the flow can be different, everything.
58
00:03:31,640 --> 00:03:35,840
So I feel like I have a better toolbox and I'm not so restricted.
59
00:03:35,840 --> 00:03:37,080
I like to be free.
60
00:03:37,080 --> 00:03:39,880
I'm free when I work with a power platform.
61
00:03:39,880 --> 00:03:41,200
>> Yeah.
62
00:03:41,200 --> 00:03:45,360
And you are, or we have, I think it's three, but
63
00:03:45,360 --> 00:03:50,040
you are specialized in model driven apps instead of Canvas apps.
64
00:03:50,040 --> 00:03:53,560
Why did you prefer model driven apps?
65
00:03:53,560 --> 00:03:56,640
And I think that's also some new.
66
00:03:56,640 --> 00:04:00,160
Generative apps, I don't know, I have forgotten the name.
67
00:04:00,160 --> 00:04:00,760
>> Yeah.
68
00:04:00,760 --> 00:04:05,960
Well, I don't, I wouldn't say like I prefer it.
69
00:04:05,960 --> 00:04:10,960
I come from that, since I've been a dynamic girl,
70
00:04:10,960 --> 00:04:15,760
of course, like model driven apps are closer to where I started.
71
00:04:15,760 --> 00:04:20,200
But it's just that in most cases where I find a process,
72
00:04:20,200 --> 00:04:26,080
I see that you get so much for free, like when you create a model driven app.
73
00:04:26,080 --> 00:04:32,600
It's such quick development and you get so much functionality that supports.
74
00:04:32,600 --> 00:04:37,840
And it's like recognizable for someone who's working with Excel or
75
00:04:37,840 --> 00:04:40,600
even outlook or any of the other Microsoft products.
76
00:04:40,600 --> 00:04:46,400
So, and then Canvas apps absolutely have their use cases.
77
00:04:46,400 --> 00:04:50,320
So, or maybe not anymore, we'll see.
78
00:04:50,320 --> 00:04:58,200
But they had and so there's absolutely use cases where I've used Canvas apps too.
79
00:04:58,200 --> 00:05:04,600
I just feel that in very many cases, the overhead becomes too much in those cases,
80
00:05:04,600 --> 00:05:07,240
where you need like a standalone Canvas app.
81
00:05:07,240 --> 00:05:14,400
Because it's still quite gruesome to configure and do complex stuff.
82
00:05:14,400 --> 00:05:20,600
A model driven app, you can have a generative page in now,
83
00:05:20,600 --> 00:05:23,680
which is like my favorite thing ever.
84
00:05:23,680 --> 00:05:29,040
This is, I just came home yesterday evening from European Power Platform Conference,
85
00:05:29,040 --> 00:05:33,760
where I had two talks and both of them had generative pages in them.
86
00:05:33,760 --> 00:05:39,200
Because I see this feature like solving so many problems.
87
00:05:39,200 --> 00:05:40,480
It's really a chameleon.
88
00:05:40,480 --> 00:05:44,320
It can, it can adapt to so many different use cases.
89
00:05:44,320 --> 00:05:48,600
So, you have the base and the quick development of a model driven app
90
00:05:48,600 --> 00:05:51,960
and all the benefits and all the navigation and search functionality
91
00:05:51,960 --> 00:05:57,240
and all of that that comes out of the box with a model driven app.
92
00:05:57,240 --> 00:06:02,960
And then you can, when you need to, extend it with a generative page
93
00:06:02,960 --> 00:06:05,400
and the UI could be whatever you want.
94
00:06:05,400 --> 00:06:14,040
It could be a multi-table wizard or it could be some type of visual graphic,
95
00:06:14,040 --> 00:06:18,040
maybe a board, a can-band board, a scheduling board,
96
00:06:18,040 --> 00:06:20,960
a calendar of some sort to display data.
97
00:06:20,960 --> 00:06:26,640
Anyway, you need to display data in a different way than a model driven app.
98
00:06:26,640 --> 00:06:30,280
Because we all know also model driven apps have a couple of downsides.
99
00:06:30,280 --> 00:06:36,000
And it's a bit clicking in some cases when you need to create a lot of records.
100
00:06:36,000 --> 00:06:42,080
So that is also a great use case for a gen page where you,
101
00:06:42,080 --> 00:06:45,960
you have like a process where you need to create a lot of records
102
00:06:45,960 --> 00:06:49,200
for example, for many, too many relationships.
103
00:06:49,200 --> 00:06:53,160
Then it would be like, I think five or six clicks for each record.
104
00:06:53,160 --> 00:06:57,160
And here you can just do it in one go with a few clicks.
105
00:06:57,160 --> 00:07:00,040
You just change the UI with a gen page.
106
00:07:00,040 --> 00:07:05,560
So the combination of all the benefits of model driven apps
107
00:07:05,560 --> 00:07:11,320
plus a gen page is like the future of apps.
108
00:07:11,320 --> 00:07:12,640
No, I don't know.
109
00:07:12,640 --> 00:07:15,040
Yeah, I'm really excited about it.
110
00:07:15,040 --> 00:07:21,120
When we start with an or with opening power apps,
111
00:07:21,120 --> 00:07:23,600
is there any requirements or to do it?
112
00:07:23,600 --> 00:07:29,120
I have to do before I show start the app studio.
113
00:07:29,120 --> 00:07:32,160
With with gen pages, particularly you mean?
114
00:07:32,160 --> 00:07:33,320
Yeah.
115
00:07:33,320 --> 00:07:36,880
So you need to you need to have a model driven app.
116
00:07:36,880 --> 00:07:44,000
You configure it from from the app builder like that the model driven app designer
117
00:07:44,000 --> 00:07:44,840
probably is called.
118
00:07:44,840 --> 00:07:45,200
I don't know.
119
00:07:45,200 --> 00:07:48,520
They change the name so many times.
120
00:07:48,520 --> 00:07:51,680
But then there's two ways of creating these.
121
00:07:51,680 --> 00:07:56,040
And they are two very different experiences.
122
00:07:56,040 --> 00:08:01,520
So you have the UI in the make report all that creates like
123
00:08:01,520 --> 00:08:08,520
they give you a prompt box where you can describe the page that you want.
124
00:08:08,520 --> 00:08:11,240
The UI, you want the problem you want to solve.
125
00:08:11,240 --> 00:08:13,760
You add your data sources.
126
00:08:13,760 --> 00:08:20,640
Maybe you even add an image of how you want this to look what you want this to look like.
127
00:08:20,640 --> 00:08:24,320
And then it generates a page for you.
128
00:08:24,320 --> 00:08:34,680
The thing here is that that way of doing it in the UI is only GA in the United States,
129
00:08:34,680 --> 00:08:39,880
UK, Australia, maybe some more English speaking language country.
130
00:08:39,880 --> 00:08:45,280
But mostly those those countries and it's in preview.
131
00:08:45,280 --> 00:08:50,800
For example, Sweden and Finland and Germany and all of those other countries.
132
00:08:50,800 --> 00:08:59,280
It only gives you the option to use model GPT 4.1.
133
00:08:59,280 --> 00:09:05,280
And that is a retired model from OpenAI.
134
00:09:05,280 --> 00:09:11,040
And as you can hear by the number of it, it's not the newest model.
135
00:09:11,040 --> 00:09:16,720
So I would say that reflects a lot in the results too.
136
00:09:16,720 --> 00:09:22,800
And what happens when you do this is also you get a little chat box and an agent you can
137
00:09:22,800 --> 00:09:27,960
continue to prompt and develop your page in.
138
00:09:27,960 --> 00:09:34,520
And that every time you add a prompt or like add it to change something, it regenerates the
139
00:09:34,520 --> 00:09:35,680
whole page.
140
00:09:35,680 --> 00:09:39,360
So I don't say it's so efficient.
141
00:09:39,360 --> 00:09:43,000
So there's and then there's some quirky bugs in the UI too.
142
00:09:43,000 --> 00:09:50,680
So like I would say the UI way of doing it, the maker portal way of doing it, it's mostly
143
00:09:50,680 --> 00:09:58,200
for trying it out and maybe at most doing a proof of concept maybe.
144
00:09:58,200 --> 00:10:07,160
But if you're if you really want to do it the solid way, I would go into any coding tool
145
00:10:07,160 --> 00:10:08,240
of your choice.
146
00:10:08,240 --> 00:10:18,360
I use visual studio VS code and GitHub code pilot in combination.
147
00:10:18,360 --> 00:10:21,800
And to emphasize here, I'm not a developer.
148
00:10:21,800 --> 00:10:23,800
I do not have a developer background.
149
00:10:23,800 --> 00:10:27,120
This is not necessary for using these tools.
150
00:10:27,120 --> 00:10:34,280
But what you get here is that Microsoft has released a skill, a plugin base that you
151
00:10:34,280 --> 00:10:41,080
can install in VS code that a skill for gen pages.
152
00:10:41,080 --> 00:10:48,840
And it's really well written and that means you can pretty much go and prompt your page
153
00:10:48,840 --> 00:10:52,920
where you actually have the choice to choose your AI model.
154
00:10:52,920 --> 00:10:57,040
The more complex page the higher the model is required.
155
00:10:57,040 --> 00:11:04,200
So you can go for for cloud any type you want depending on your want to do results.
156
00:11:04,200 --> 00:11:09,240
The skill also include like a planning stage which I kind of like.
157
00:11:09,240 --> 00:11:21,320
So you talk with your agent in VS code and like discuss the layout and it creates a TML
158
00:11:21,320 --> 00:11:26,920
wire frames which you can click through and like yes, okay, now we're on onto something.
159
00:11:26,920 --> 00:11:32,600
This is what I want or you can before it starts building.
160
00:11:32,600 --> 00:11:37,080
So that will also be before except for it being more efficient.
161
00:11:37,080 --> 00:11:43,760
It will also save you a credit since most coding tools now or at least get up code pilot
162
00:11:43,760 --> 00:11:47,080
is a credit based.
163
00:11:47,080 --> 00:11:53,960
So so I and if you go to the UI and you do the prompt in the make report for gen page,
164
00:11:53,960 --> 00:12:01,520
even the UI has this little box at the top level like do you want to enable more AI models?
165
00:12:01,520 --> 00:12:04,560
The way to go is with via coding tools.
166
00:12:04,560 --> 00:12:09,080
So they're like really pushing in that direction, but I would say like the biggest benefit here
167
00:12:09,080 --> 00:12:17,320
is that SGA world wide if you go with if you make it through the coding tools.
168
00:12:17,320 --> 00:12:22,520
So I can implement one and it would be supported and all of that that way.
169
00:12:22,520 --> 00:12:28,160
And I would get a so much better result than doing it in the make report.
170
00:12:28,160 --> 00:12:33,920
And before I start how how important is to think about I don't know the business value,
171
00:12:33,920 --> 00:12:40,880
the strategy, what tips can you give here and before it's come to the first prompt?
172
00:12:40,880 --> 00:12:50,920
>> So first of all, you have to identify that a model driven app,
173
00:12:50,920 --> 00:12:55,360
the UI of a model driven app is not enough for your use case, right?
174
00:12:56,640 --> 00:13:00,640
But you need to identify that it also belongs in this app somehow.
175
00:13:00,640 --> 00:13:06,400
Like you still want to work with the same data that is in the rest of the app.
176
00:13:06,400 --> 00:13:14,160
It has to be the same target audience like user group that is actually using this data
177
00:13:14,160 --> 00:13:21,120
to be able to or to take the decision to make a gen page out of it.
178
00:13:21,120 --> 00:13:30,800
So absolutely, you need to look at the business side and see if there's a proper use case
179
00:13:30,800 --> 00:13:32,560
in the app that you're talking about.
180
00:13:32,560 --> 00:13:41,520
But we are also shifting because like return of investment now when you build it this way,
181
00:13:41,520 --> 00:13:48,080
because it's also like a million times quicker to build a very complex gen page
182
00:13:48,080 --> 00:13:53,920
compared to like a custom page, which is still low code,
183
00:13:53,920 --> 00:13:57,680
but like it could take a lot of time to create a custom page with power effects
184
00:13:57,680 --> 00:14:00,880
and all the configuration that you have to do for it to work.
185
00:14:00,880 --> 00:14:06,480
And here is like a few prompts here and there and UK work with other stuff
186
00:14:06,480 --> 00:14:09,440
in between while the agent is working too.
187
00:14:09,440 --> 00:14:17,440
So I would say like before maybe we made decisions not to implement a custom page
188
00:14:17,440 --> 00:14:22,320
or a complex UI because like, oh, it would be too much to maintain
189
00:14:22,320 --> 00:14:26,960
and to extend and continue to develop.
190
00:14:26,960 --> 00:14:34,000
And we would give the users unrealistic expectations of the UI
191
00:14:34,000 --> 00:14:36,320
and other parts and stuff like that.
192
00:14:36,320 --> 00:14:43,040
I don't think we have to be equally afraid of that now as we maybe were before.
193
00:14:43,040 --> 00:14:49,840
So like maybe before I was an architect that pulled the brakes a bit more than I will do now.
194
00:14:49,840 --> 00:14:59,120
I would say no a lot more before than I will now because of the speed of development of this feature.
195
00:14:59,120 --> 00:15:06,160
And when what makes a good prompt or what separates a good prompt for a bad prompt
196
00:15:06,160 --> 00:15:13,600
to build a page especially when we think about building an app prompt?
197
00:15:13,600 --> 00:15:21,440
That's a very good question and it kind of depends on where you start.
198
00:15:21,440 --> 00:15:26,880
So and what instructions your agent have that you are using.
199
00:15:26,880 --> 00:15:35,040
In the UI I would use as little description for the interface
200
00:15:35,040 --> 00:15:37,600
like how you wanted to look as possible.
201
00:15:37,600 --> 00:15:41,920
I rather just attach an image to make it more similar.
202
00:15:41,920 --> 00:15:49,280
But to be honest my biggest tip here is to go to any of your favorite tool.
203
00:15:49,280 --> 00:15:57,680
If it's co-pilot chat or if it's chat GPT or whatever, describe shortly what you want to create
204
00:15:57,680 --> 00:16:01,040
and have them help you refine the prompt.
205
00:16:01,040 --> 00:16:07,440
Because they are so much better at structuring a prompt and making more sensible.
206
00:16:07,440 --> 00:16:15,440
And also efficient if you say that the bigger the prompt the more tokens it will use.
207
00:16:15,440 --> 00:16:21,840
It will also be more efficient especially for someone who's not have English as a native language.
208
00:16:21,840 --> 00:16:30,240
Maybe I have a tendency to use a bit more words than a native English speaker and then your AI.
209
00:16:30,240 --> 00:16:35,040
How can help you make it less?
210
00:16:35,040 --> 00:16:36,400
So what would you say?
211
00:16:36,400 --> 00:16:44,240
What do you think one big prompt is better or an iterative prompting process?
212
00:16:44,240 --> 00:16:50,640
So the context window if it's a small prompt to start with,
213
00:16:50,640 --> 00:16:54,880
the context window will be wider.
214
00:16:54,880 --> 00:17:01,680
So as many details as you have I would try to add to my first prompt.
215
00:17:01,680 --> 00:17:05,680
But if you go with the code in two way, which is my recommended way,
216
00:17:05,680 --> 00:17:11,680
like since it doesn't start building until you tell it to start building,
217
00:17:11,680 --> 00:17:20,480
I wouldn't be too afraid of having like a conversation afterwards because there's so much things that you realize when you get there,
218
00:17:20,480 --> 00:17:27,280
the wireframes like, oh, that that would not at all what I meant or what what I want or oh,
219
00:17:27,280 --> 00:17:33,280
I forgot that I should be able to love a lot in this page too.
220
00:17:33,280 --> 00:17:37,280
So so yeah, I think you could go either way,
221
00:17:37,280 --> 00:17:45,280
but start with as many details as you have so you don't have to jump through those hoops afterwards.
222
00:17:45,280 --> 00:17:54,480
And when I do this app, I think a little bit about yeah, main intents, how can other understand the app?
223
00:17:54,480 --> 00:18:00,080
I don't know someone we try out or leave the company from other reason.
224
00:18:00,080 --> 00:18:04,080
How can we have there on this IE generated pages?
225
00:18:04,080 --> 00:18:05,680
How we have the overview?
226
00:18:05,680 --> 00:18:09,680
What's happened in the app?
227
00:18:09,680 --> 00:18:15,280
And what's the question, Sir?
228
00:18:15,280 --> 00:18:20,080
Yeah, I think a little bit like governance or development question,
229
00:18:20,080 --> 00:18:25,280
but it's someone have make a lot of citizens develop, no, everyone can make apps,
230
00:18:25,280 --> 00:18:31,280
but someone leaves the company or or it's sick for a while and I have to change something on the app.
231
00:18:31,280 --> 00:18:37,680
How can I look into the app and get to understand how an AI generated app works?
232
00:18:37,680 --> 00:18:40,480
What's to your chips here?
233
00:18:40,480 --> 00:18:48,080
Well, hopefully the AI has helped you design it, so it's kind of self explanatory.
234
00:18:48,080 --> 00:18:59,680
But you can also add so it again depends on what type of agent you're using in your coding tool.
235
00:18:59,680 --> 00:19:16,080
So I have a consultant agent that helps me discuss a lot about the process itself first
236
00:19:16,080 --> 00:19:21,280
and it asks me questions about the process, it finds gap in the process,
237
00:19:21,280 --> 00:19:24,080
and it asks like, oh, so how do you want to do with this part?
238
00:19:24,080 --> 00:19:27,680
And I was like, wow, that's out of school or all of that.
239
00:19:27,680 --> 00:19:37,280
I can't, since I have all this conversation and I have therefore a plan and a process a plan to out,
240
00:19:37,280 --> 00:19:42,880
I have half of the documentation already done with that agent.
241
00:19:42,880 --> 00:19:50,080
It already made me mermaid diagram, it already made me the wire frames and all of those parts.
242
00:19:50,080 --> 00:19:56,480
It even makes me a plan for implementation before design.
243
00:19:56,480 --> 00:20:06,080
So all of that you can give to a new developer, but you can also ask it to spit out like user manual based on this.
244
00:20:06,080 --> 00:20:19,280
And you can add a playwright, MCP to grab a bunch of screenshots from the, from the, the wallpaper to yourself.
245
00:20:19,280 --> 00:20:27,280
And then it will add that to you, whatever instruction manual you want.
246
00:20:27,280 --> 00:20:33,280
Yeah, and in the last time, I think that's also upcoming topic.
247
00:20:33,280 --> 00:20:41,680
It's especially even we call about a I generated apps that's data was what role do data was play?
248
00:20:41,680 --> 00:20:46,880
You mean data sources or what?
249
00:20:46,880 --> 00:20:48,880
Like the quality of data.
250
00:20:48,880 --> 00:20:54,880
The data verse is all talk about it, especially when it's come to ads.
251
00:20:54,880 --> 00:20:58,880
Why is this such an important topic actually?
252
00:20:58,880 --> 00:21:13,880
So sorry, I just got lost track here, but one thing that is coming by the way that I saw for Gen pages that I saw this week is that right now Gen pages only works with the data verse.
253
00:21:13,880 --> 00:21:17,880
So you can only extend a model driven app.
254
00:21:17,880 --> 00:21:23,880
The good thing about custom pages is that you can actually infuse other data sources in a custom page.
255
00:21:23,880 --> 00:21:31,880
So you can have other data sources than data verse in your model, or an app big advantage.
256
00:21:31,880 --> 00:21:41,880
Leon announced at if you receive that it's coming connectors are coming for for generative pages too.
257
00:21:41,880 --> 00:21:45,880
So even in July, so it's July now, right?
258
00:21:45,880 --> 00:21:55,880
So very soon, hopefully, we will see the possibility to have different data sources in in Gen pages too.
259
00:21:55,880 --> 00:21:57,880
And I lost your question.
260
00:21:57,880 --> 00:21:59,880
I got sidetracked.
261
00:21:59,880 --> 00:22:01,880
It's really interesting.
262
00:22:01,880 --> 00:22:05,880
I love this that's better than I ask.
263
00:22:05,880 --> 00:22:15,880
So that's actually the you only have data worse on and your internal, I don't know.
264
00:22:15,880 --> 00:22:27,880
Microsoft fabric or something as source and now in the future, we get really extendable sources from from from outside or.
265
00:22:27,880 --> 00:22:40,880
Exactly. So I don't know if it's so I just saw this like on an overview slide. So I don't know if it's through connectors is just a few connectors.
266
00:22:40,880 --> 00:22:43,880
I don't know what data sources yet, but I just saw it.
267
00:22:43,880 --> 00:22:48,880
It's coming. So Microsoft has definitely realized that.
268
00:22:48,880 --> 00:22:59,880
You need a way to be able to extend your model to an app with more data than just data verse in this is the way way to do it going forward I would get.
269
00:22:59,880 --> 00:23:01,880
So I'm really excited about that.
270
00:23:01,880 --> 00:23:08,880
Another thing that they announced at this conference is that.
271
00:23:08,880 --> 00:23:20,880
That's been a huge ask for everyone working with custom pages for for ages is to be able to embedded on a form or a tab like you can do with a camera's app.
272
00:23:20,880 --> 00:23:28,880
Because that way with cameras up is so clunky and the response of this is quite horrible and.
273
00:23:28,880 --> 00:23:48,880
Unfortunately had been had forced to do that a couple times and it's very messy. But that is coming with generative pages also very soon that you are supposed to be able to enable it in a tab or even section it said.
274
00:23:48,880 --> 00:24:04,880
I'm really intrigued how this will work like not only responsive and that's wise, but also performance wise because right now model driven forms are quite quick to load.
275
00:24:04,880 --> 00:24:28,880
And how will having a gen page embedded effect that and what does this mean compared to PCF that are also UI that you can embed like on a form a custom control that will change the UI will this complete the replace the need for PC apps or.
276
00:24:28,880 --> 00:24:36,880
So I don't know where we going but I'm very excited to see where where this takes us.
277
00:24:36,880 --> 00:24:46,880
Yeah, I think the PCF is the power apps component component framework. I think this is really cool stuff because you have all these ready made.
278
00:24:46,880 --> 00:24:56,880
Yeah, I don't know things like the I don't know I love the barcode scanner, but I hope they're there comes more in future.
279
00:24:56,880 --> 00:25:12,880
What make make did it for you make sense build forms instead of apps watch or I every use apps instant or forms.
280
00:25:12,880 --> 00:25:41,880
I mean, I meant like the the forms of a record in a model driven app. That is what you can now or hopefully soon embed a gen page on so it would pretty much be as a PCF partly where you can extend the UI with both different data sources maybe like dress up a record with information from.
281
00:25:41,880 --> 00:25:54,880
From your ERP or something or you could have like a mini wizard embedded on the form I don't know yet I haven't seen it but I'm I'm excited.
282
00:25:54,880 --> 00:26:04,880
Yeah, okay, I really yeah, I don't know it's it's really new or the PC PCF or is it.
283
00:26:04,880 --> 00:26:24,880
So it's I would say like it's almost a replacement of PCF it's like the it's gen pages embedded in in a model driven form that's that's the key like here but that is coming in July and I haven't seen it yet so I don't know how it's going to work.
284
00:26:24,880 --> 00:26:38,880
And then let's think about one one enterprise topic is also a sink governance. Why think you many organizations try all this power platform governance.
285
00:26:38,880 --> 00:26:53,880
Because with the the rapid development right now how quickly you can create components in the power platform. The more important it becomes to govern everything that is created right when you do.
286
00:26:53,880 --> 00:27:16,880
We need structure to make sure that our agents is behaving as we want them to behave that our apps are performing in our extendable and have an owner and all of that I see like we will have before we were maybe more limited about how many apps can we create in a year.
287
00:27:16,880 --> 00:27:45,880
That will change now to how like how many apps can we adopt because the creation part will will not be the hardest part anymore it will be more about adoption continuous development governance make sure that it keeps in their guardrails and with the constraints that is necessary and I think we need to talk more about the processes itself and why why.
288
00:27:45,880 --> 00:28:14,880
Why we need an app or why we need an agent or or if we should rethink the whole process and need are we just I think it was don as our car I heard at a key note that we're comparing AI to to cars and horses like we don't use AI to to make faster horses.
289
00:28:14,880 --> 00:28:43,880
We have to make the horses that are our transport for now faster we have to rethink transport overall and therefore cars was created and we are in the same with with like business support business systems that we maybe need to any I hear that we need to recreate the thinking of how we solve problems can complete.
290
00:28:43,880 --> 00:28:49,880
We don't want to solve the same problems just with AI if that makes sense.
291
00:28:49,880 --> 00:28:57,880
I don't think I'm a good storyteller as Donna she made a lot more sense when she said.
292
00:28:57,880 --> 00:29:20,880
But a lot of companies are think years before we have these data driven companies AI driven companies but now I think one of the other top of the keys is a city developer company how should come and you think about citizen development when they start with it what tips can you give here.
293
00:29:20,880 --> 00:29:43,880
That's a good question I think we were in the same challenges as we've been ever since low code entered our our sphere that the people are going to want to build stuff themselves and and really maybe they are the best at doing it because they know the process really well itself.
294
00:29:43,880 --> 00:30:08,880
So I think it about enabling those people who want to do it themselves and teaching them skills that they need to be aware of like security issues data access all of the things you need to consider when you you design an app or an agent that we should give them the tools and information to help them with that.
295
00:30:08,880 --> 00:30:10,880
That's that's pretty much it.
296
00:30:10,880 --> 00:30:14,880
Okay, I think it's a good tip.
297
00:30:14,880 --> 00:30:33,880
Also, what can you say a lot of people I think there are yeah, but hearing a little bit about starting develop their own own app which can you give people who will start or think about to start with citizen development.
298
00:30:33,880 --> 00:31:02,880
I would say I would say just just make sure to design the process first don't like previously 10 14 years ago when I started out we were kind of limited to pushing our processes into existing systems because it would
299
00:31:02,880 --> 00:31:14,880
be too expensive to change parts of the systems some some things we just had to like oh this is how it is this is we have to buy into this.
300
00:31:14,880 --> 00:31:30,880
All of that now especially with AI we really can forget about the tool that will get us there we can really really talk about the process itself first hand.
301
00:31:30,880 --> 00:31:48,880
So I would really like spend time in updating your skill on designing the process and then like the making part girl that that's that's easy that will be that will be the smallest.
302
00:31:48,880 --> 00:32:17,880
Yeah, I think one one yeah controversial topics are also in companies that's the sketch is the platform power platform center of excellence what advice can you give here or what's your experience with center of excellence well I think they I think I don't know the proper term but I think they retired it a couple of months ago or a month ago or something that they will not continue to develop that because the app.
303
00:32:17,880 --> 00:32:46,880
Because the admin portal is the getting better better and have most of the features that they create the the kit center excellence kit have because it was just always a starting point right and now the the admin portal has most of that information and gives you the governance that you need and if not you can build you can maybe build some of it yourself.
304
00:32:46,880 --> 00:33:06,880
So I wouldn't implement that now if you have implemented already continue to use it for as long as it makes sense but look at the features in the admin portal to see how redundant it is because it was a hassle to maintain a center excellence kit.
305
00:33:06,880 --> 00:33:32,880
Yeah, we have now so much new I say words or or a bus but a little bit yeah losing sometimes the overview we have now generative ages custom pages PCS what what or how did you have can have the overview of all what they're there are coming you how did you handle this.
306
00:33:32,880 --> 00:34:01,880
So it's I think it's an impossible task and I don't think I have the the overview or the whole picture I'm I'm trying to to stay stay current just as everyone else and try try to understand I go to conferences I read blogs I look at linkedin I look at like the release notes and and all of that I'm still probably missing a lot of stuff.
307
00:34:01,880 --> 00:34:29,880
So I would say like I use the community a lot and I use my colleagues a lot because and when everything is happening so fast in our platform no one can know at all that means that I always want to make sure to check my solution that I'm designing with one or two or even three other people to make sure that there's no other way no better way to do it.
308
00:34:29,880 --> 00:34:37,880
Yeah, like cross check and report everything because it's impossible to know everything yourself.
309
00:34:37,880 --> 00:34:47,880
Did you see a big picture Michael's have had with with with power pages overall.
310
00:34:47,880 --> 00:34:53,880
Power pages I know nothing about not one thing or.
311
00:34:53,880 --> 00:35:15,880
When we are yeah I think when we build these these all all these is nice stuff and you have say a community and so on and you are also a founder of the power platform community Sweden can you tell a little bit about this initiative.
312
00:35:15,880 --> 00:35:32,880
Yeah it's I started before the pandemic I thought like because I heard my my international friends I was going to international conference speaking back then too and they have like user groups.
313
00:35:32,880 --> 00:35:59,880
The like local communities where they had a full day of learning or or something so I find a user group in Sweden and I went to that a couple of times and spoke at that but then I was like it's so you sure focused in Sweden in in UK it seems to be a lot of like makers and users combined but in Sweden it was like so you sure focused only like the customer side of it and I was like but what.
314
00:35:59,880 --> 00:36:26,880
There need to be like a forum for for everyone who's building everyone who's maker if you're a consultant or you work for an end customer it doesn't matter but like where are the maker group so I founded that myself so I reach out to old colleagues and people only then it was like i'm trying to do this is this even of interest and I started organizing these events so it's like.
315
00:36:26,880 --> 00:36:49,880
I want to make it very easy to stop by and low low bar for people to come so we do like after work style so after work is done from 5 30 to 8 30 we have two sessions we are at different partners every time that like vice peach says and beers so we watch a couple sessions.
316
00:36:49,880 --> 00:37:18,880
Someone is presenting something fun and learning we geek out about power platform we have a lot of there you get to know people and it work with the people locally and that really took off so so well so now in my Stockholm events I I always have a wait list like always is always fully booked and I have a long queue of people that wants to come in but can't get a seat.
317
00:37:18,880 --> 00:37:24,880
Maybe I should try to have more events but I don't really have time for.
318
00:37:24,880 --> 00:37:47,880
But then we also expanded to to Malamana and Gothenburg which is to the two other main cities in Sweden so I have like a group of people who started organizing the south south branch i'm only supporting with like the website and some tips on speakers.
319
00:37:47,880 --> 00:38:11,880
Mostly they're running by themselves and Gothenburg has just started off trying to support them too as much as I can but like it's franchise here okay so they're they're kind of on their own and I'm just helping a bit but it's really fun to see that it's growing and it's the people really want to go and like the concept of it.
320
00:38:11,880 --> 00:38:22,880
And yeah you are also speaker what did you learn from yeah from speaking on on other events and also getting to other events.
321
00:38:22,880 --> 00:38:40,880
I I love that I get to really deep dive in the subject before I go on stage like I really need to figure out all the little details about the feature that I'm talking about so it's like an opportunity for myself to force me to.
322
00:38:40,880 --> 00:38:56,880
Really go deep in something that that is one of the main advantages of going on stage and then it's also extremely fun and to talk about something that you're nerdy about and like you.
323
00:38:56,880 --> 00:38:58,880
Yeah.
324
00:38:58,880 --> 00:39:08,880
Yeah if you say a feature in the deep dive is there one feature you say oh the visits yeah it's it's underrated actually.
325
00:39:08,880 --> 00:39:25,880
Underrated feature well I would almost say that Jen Peeces Arsenal underrated I don't I I talk very loudly about it but I don't think it's talked about enough.
326
00:39:25,880 --> 00:39:33,880
I yeah no I think that is my biggest one that I would push even more if I could.
327
00:39:33,880 --> 00:39:47,880
And how what tips can you give like I think someone is new and tech and also likes to become an architect like you how actual they start or he.
328
00:39:47,880 --> 00:40:01,880
I think that is the biggest question of our of our occupation right now like how do we get people that starting out on this train because.
329
00:40:01,880 --> 00:40:10,880
AI is actually replacing some of the junior tasks that like new people in tech was doing.
330
00:40:10,880 --> 00:40:30,880
So again I would it seems like a red thread through this conversation but I would would ask them to focus on the the business side of it like the designing processes sketching out flows process flows and diagrams and all of those skills.
331
00:40:30,880 --> 00:40:39,880
Because those are the ones that's going to matter in the future not learning where to click in the in the make report all.
332
00:40:39,880 --> 00:40:54,880
And what will you do when Microsoft came and say hey Sarah you get all the resources all the money the needs you can develop one feature which which will you develop.
333
00:40:54,880 --> 00:41:08,880
Oh I am no clue I am too realistic I have no clue what I would I would do.
334
00:41:08,880 --> 00:41:15,880
Okay you think it's not realistic that such an ideal you're actually goes tomorrow breaks.
335
00:41:15,880 --> 00:41:30,880
Yeah so I have in every podcast I have a quick fire out short questions I say then you give a short short answer so let's start simple coffee.
336
00:41:30,880 --> 00:41:40,880
I did you think or tea through the event of it coffee early bird or night all both.
337
00:41:40,880 --> 00:41:45,880
You know I'm favorite Microsoft build announcement.
338
00:41:45,880 --> 00:41:50,880
Wow I know clue.
339
00:41:50,880 --> 00:41:58,880
Yeah what is your favorite the conference.
340
00:41:58,880 --> 00:42:07,880
No I can't say I go to so many and I would piss people off no matter my answer so I can't say.
341
00:42:07,880 --> 00:42:13,880
But is there any book you can yeah you say everyone should read.
342
00:42:13,880 --> 00:42:17,880
Hitch hiker's guide to the galaxy is my favorite book.
343
00:42:17,880 --> 00:42:22,880
That is really cool.
344
00:42:22,880 --> 00:42:27,880
What was your biggest lesson from your career.
345
00:42:27,880 --> 00:42:30,880
Don't be afraid to stand out.
346
00:42:30,880 --> 00:42:38,880
And one prediction for Microsoft AI in the next years.
347
00:42:38,880 --> 00:42:47,880
No okay so I probably be wrong so no.
348
00:42:47,880 --> 00:42:57,880
So one when we think a little bit about the near feature food future we have that's throughout today.
349
00:42:57,880 --> 00:43:12,880
And what is some and I don't know if you're out but you can say with all expose the name what's the next project you will start.
350
00:43:12,880 --> 00:43:21,880
I'm not so focused on projects right now because I'm going on Nordic vacation and Nordic vacation is like you know several weeks.
351
00:43:21,880 --> 00:43:33,880
So my focus right now is to finish up everything that I'm doing like all the conversation I have with Microsoft PMs about features and all my customers and all of that.
352
00:43:33,880 --> 00:43:47,880
I'm just trying to wrap stuff up so I can delete teams and outlook for five weeks and just be in my summer cottage and read books.
353
00:43:47,880 --> 00:43:54,880
Oh yeah it was was really great.
354
00:43:54,880 --> 00:43:59,880
And we have all we have so many topics.
355
00:43:59,880 --> 00:44:12,880
And then yeah my last question is what should people take away from this podcast when they remain meaning only one one thing.
356
00:44:12,880 --> 00:44:19,880
So the same pages are awesome and you should try it out but only the code way.
357
00:44:19,880 --> 00:44:32,880
Yeah okay so yeah Sarah thank you for being here this was yeah in info inside for conversation and I have run the Nordic holidays along the then the drones.
358
00:44:32,880 --> 00:44:43,880
I'll explore everything from enterprise architecture governance model there's an apps generatives apps AI cost of pages PCS and what's really take to build the stable business solutions in the power platform.
359
00:44:43,880 --> 00:44:48,880
So thank you for much for sharing your experience and your practical insights with us.
360
00:44:48,880 --> 00:44:55,880
And yeah I hope we see us also on conference also on and yeah have a nice day. Thank you so much.
361
00:44:55,880 --> 00:44:58,880
Thank you for having me.
362
00:44:58,880 --> 00:44:59,880
Bye.
363
00:44:59,880 --> 00:45:00,880
Bye.
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.
AI and Application Architect
I am an Application Architect working with Power Platform and AI, focused on translating real business needs into sustainable, well-governed solutions. My work sits at the intersection of business and technology, where clarity, structure, and long-term value matter more than quick fixes.
I lead requirement analysis, workshops, and solution design in complex environments, often with high demands on governance, data quality, and scalability.
I am an international speaker, presenting at events around the world, and I an the founder of Power Platform Community Sweden and run the Stockholm-branch, where I work to strengthen knowledge-sharing and community within the Power Platform ecosystem.