Turn your real-world experience into part of the show.

Microsoft Data Podcast – Analytics, Fabric & Data Governance Episodes

Modern Microsoft architectures are increasingly data-driven, spanning analytics, reporting, automation, and AI workloads. Data Talk focuses on data platforms and governance within the Microsoft ecosystem, including Microsoft Fabric, analytics services, and enterprise data architectures.

Episodes cover topics such as data modeling, pipelines, governance models, security boundaries, and integration with Microsoft 365 and Azure. We explore how data flows across services, where governance breaks down, and why unmanaged growth creates long-term risk.

Data Talk emphasizes understanding data systems as operational platforms rather than isolated analytics tools. Discussions include identity and access models, compliance considerations, data residency, and the interaction between AI-driven services and enterprise data.

This category is intended for data architects, engineers, and IT leaders responsible for designing and governing Microsoft data platforms. If you need clarity on how Microsoft data services behave at scale, Data Talk provides grounded, experience-based insight.

Power by: https://datascience.m365.fm
Nov. 18, 2025

Stop DAX UDF Timeouts: Materialize with ADDCOLUMNS Instead

You thought VAL vs EXPR was harmless syntax sugar. In reality, it decides when your metrics are evaluated, which means the same function, same arguments can return a completely different “truth” — clean visuals, perfect totals, and numbers that are confidently wrong.
Guest: Mirko Peters
Listen to the Episode
Nov. 17, 2025

How to Keep 3D Models From Bypassing Your Fabric Governance Rules

Think spreadsheets are chaos? Cute. In this episode we stress-test Microsoft Fabric against the worst data you own: photorealistic 3D assets and full-fidelity digital twins. We break down why a single “file” is actually a sprawling supply chain of scans, meshes, textures, physics, and licenses — and how one sloppy ZIP export can turn into a global compliance nightmare.You’ll learn how Fabric turns governance from theater into an always-on safety system: Entra ID–backed identity, object-level security, lineage that behaves like a black-box recorder, rights-as-code, and streaming with signed tokens instead of random file copies. We walk through real workflows for artists, simulation engineers, and robotics teams, and show how to version twins so “latest” stops being a ticking time bomb.If you’re still trusting folders, shared drives, and good intentions to protect multi-gigabyte 3D assets, this episode is your wake-up call. Hit play to see why, if your governance can hold a 1:1 …
Guest: Mirko Peters
Listen to the Episode
Nov. 15, 2025

Fix This Data Bottleneck Before You Buy More NVIDIA Blackwell GPUs

Your GPUs aren’t the problem. Your data fabric is.In this episode, we unpack why “AI-ready” on top of 2013-era plumbing is quietly lighting your cloud bill on fire—and how Azure plus NVIDIA Blackwell flips the equation. Think thousands of GPUs acting like one giant brain, NVLink and InfiniBand collapsing latency into microseconds, and Microsoft Fabric finally feeding models at the speed they can actually consume data.We break down the Grace-Blackwell superchip, ND GB200 v6 rack-scale VMs, liquid-cooled zero-water-waste data centers, and what “35x inference throughput” really means for your roadmap, not just your slide deck. Then we go straight into the uncomfortable truth: once you fix hardware, your pipelines, governance, and ingestion become the real chokepoints.If you want to cut training cycles from weeks to days, slash dollars per token, and make trillion-parameter scale feel boringly normal, this is your blueprint.Listen in before your “modern” stack becomes the …
Guest: Mirko Peters
Listen to the Episode
Nov. 8, 2025

Users see wrong data in Power Apps? Here’s why SharePoint fails

In this episode of the M365 Show, we unpack the most common Power Apps mistake in the Microsoft 365 ecosystem: assuming SharePoint is a “free database.” We break down why SharePoint lists look like a database, but architecturally behave like a file-centric collaboration layer—not a transactional data engine. From delegation limits to the infamous 2,000-record wall, we explain how apps built on SharePoint scale beautifully for a month… then collapse under load, concurrency, throttling and lookup chains. If you’ve ever heard “we’ll just use SharePoint for now,” this episode will save future performance pain. Learn what a real database is, how Dataverse and SQL handle indexing + relationships correctly, and why treating SharePoint like SQL is the fastest way to kill a Power App.
Guest: Mirko Peters
Listen to the Episode
Nov. 6, 2025

Why Your Fabric Data Warehouse Is Still Just a CSV Graveyard

You paid for AI synapses — but you’re parking fossils. Fabric was engineered for real-time intelligence, semantic relationships, live context, and AI agents that reason. Most orgs treat it like a network share. The result? Zero cognition, zero AI, zero advantage — just expensive cold storage with better Microsoft branding.
Guest: Mirko Peters
Listen to the Episode
Nov. 5, 2025

Business Users Still Waiting Days for Reports? Fix It With Copilot Studio

Your company isn’t blocked by data—it’s blocked by syntax.Copilot Studio turns plain-English questions into governed Fabric queries, so “What was our revenue by quarter?” finally gets an instant, secure answer—no SQL, no tickets, no waiting. It’s not a chatbot; it’s a translation engine that remembers context, respects permissions, and makes your warehouse talk back like a smart analyst. The real bottleneck has never been tables—it’s been language. This is the fix.
Guest: Mirko Peters
Listen to the Episode
Nov. 5, 2025

Your Power BI Model Looks Fine but Your Queries Are Breaking It

Power BI is lying to you.Those neat, orderly “Applied Steps” you trust are a façade.Power Query is secretly rearranging, deferring, collapsing and sometimes ignoring the way you wrote your transform steps — which is why your filters don’t filter, your merges get weird, and refreshes suddenly go from 20 seconds to 10 minutes.The only way to stop getting blindsided is to understand the invisible engine + query folding… because THAT is the real execution order — not the thing you see in the pane.
Guest: Mirko Peters
Listen to the Episode
Nov. 4, 2025

Copilot shows wrong numbers? Fix your Fabric data model now

Copilot didn’t hallucinate — you hallucinated first.Your schema lied → Fabric believed it → Copilot repeated it with confidence.Bad Bronze → leaky Silver → fake Gold = executive decisions built on fiction.Fix the Medallion discipline + fix the semantic layer — or keep paying for an AI that politely invents your reality.
Guest: Mirko Peters
Listen to the Episode
Nov. 4, 2025

Why Your Power BI Dashboards Don’t Change Any Decisions

Power BI isn’t failing because of visuals — it’s failing because nobody plans. 60–70% of BI projects become expensive wallpaper. This episode exposes the 3 discipline steps that separate strategic intelligence from dashboard vanity — why most organizations confuse activity for progress, and how to finally stop the “successful failure” of Power BI.
Guest: Mirko Peters
Listen to the Episode
Nov. 3, 2025

Power BI Governance 101: Build One Single Source of Truth

Power BI Collaboration — from Wild West → Hub-and-SpokePower BI self-service feels empowering… until every department defines “revenue” differently and no one agrees which dashboard is real. In this episode, we break down why the chaos isn’t a tooling problem — it’s an architecture problem — and how the Hub-and-Spoke model fixes it.We walk through how to create one shared semantic truth (the Hub) — with certified datasets, owners, refresh discipline, and version control — while still letting departments move fast in their own exploration spaces (the Spokes).This is the roadmap to move your analytics org from “faith-based metrics” to governed trust.
Guest: Mirko Peters
Listen to the Episode
Nov. 2, 2025

Dataverse Is Not Free: How to Avoid the Power Apps License Trap

Dataverse isn’t “included.” It’s a premium, enterprise-grade platform with costs that stack fast: licenses, capacity, environments, storage (db/file/log), and API limits. Most sticker shock comes from assumptions—thinking M365 covers Dataverse, that guests are free, or that storage is cheap.Use Dataverse when governance, security, and relational consistency truly matter. Otherwise, cheaper setups (SharePoint Lists, SQL, or hybrid) deliver ~80% of the value without premium connectors or capacity burn. The mature play: prototype on Lists/SQL, graduate to Dataverse only when scale, audit, and security demand it.Dataverse is powerful and pricey by design. Treat it as a graduation platform, not a default. Budget first, then build.
Guest: Mirko Peters
Listen to the Episode
Nov. 1, 2025

Why Your Fabric Dataflows Are Burning Compute for Nothing

Most Microsoft Fabric teams are bleeding money because they treat Dataflows Gen2 like old Power BI ETL. In Fabric, compute—not storage—is the meter, so every redundant refresh spins up clusters, reloads the same sources, and multiplies cost. The fix is architectural, not heroic CSV exports.There are only three sane patterns:Bronze / Staging — land external data once into Delta (incremental where possible). Everyone else references, no re-ingestion.Silver / Transform — centralize business logic and data quality. Build computed entities, enforce semantics, and chain flows to bronze for clean lineage.Gold / Serve — expose curated Delta to consumers (Direct Lake semantic models, shared tables). No import refreshes, no duplication, low latency.Choose based on cost, governance, and speed. Small teams: bronze + thin silver. Enterprises: full bronze/silver/gold with centralized orchestration. Mixed mode: bronze+silver for builders, minimalist gold for exec reporting. Bottom li…
Guest: Mirko Peters
Listen to the Episode
Nov. 1, 2025

Drowning in CSVs for Audits? Automate Fabric Governance with GPT‑5

Fabric didn’t fail at features—it failed at governance cohesion. Data lives in Fabric, security sits in Power BI, labels live in Purview—and they don’t natively reason about each other. That’s why audits devolve into CSV marathons and name-matching nightmares.Enter GPT-5 inside Microsoft 365 Copilot. The leap isn’t speed; it’s reasoning. GPT-5 infers relationships across Fabric lineage, Purview classifications, and Power BI security—then validates them as one logical story. Ask, “Which Lakehouse tables with PII lack Purview labels and Power BI RLS?” It reconstructs lineage, cross-checks labels, inspects roles, and returns a verified, explainable result—no manual triangulation.The payoff: governance shifts from reactive, episodic audits to predictive, continuous assurance. Costs drop (fewer human hours and mistakes), confidence rises (explanations with evidence), and scale improves (one reasoning loop supervises thousands of assets). Implement via Copilot Studio: enable GPT-5 (…
Guest: Mirko Peters
Listen to the Episode
Oct. 30, 2025

Your Power BI Numbers Are Wrong Because of Excel – Fix the Data with Copilot

Most “analysis” in Excel is disguised janitorial work: inconsistent dates, mixed data types, rogue spaces, and copy-pasted chaos that later poisons Power BI, Power Automate, and Fabric. The fix isn’t heroics—it’s Excel Copilot acting as an AI janitor that understands structure, enforces types, and prepares data for downstream systems. Two modes matter: Chat (diagnose, explain) and App Skills (actually change the sheet). The practical playbook: 1) Normalize Everything (types, dates, casing, spaces, IDs), 2) Validate & Flag Outliers (rules for missing/absurd values, deviation checks), 3) Transform for Integration (summaries, tidy tables, headers for Power BI/Flows). Once clean, Copilot upgrades sheets from neat to smart—sentiment tagging, segmentations, “Think Deeper” diagnostics—then hands results straight to Power BI/Power Automate without brittle CSV shuffles. The manual phase is over: treat Copilot as a preprocessor at data intake, codify cleanup as prompts, and promote analysts fro…
Guest: Mirko Peters
Listen to the Episode
Oct. 23, 2025

SharePoint Governance Fail: Why Your Power Apps Are Out of Control

SharePoint Lists feel “free” and familiar, so teams prototype apps on them—and then accidentally build departments on sand. As lists multiply, you get broken lookups, 5k-item throttling, rogue permissions, attachment bloat, and schema drift. Governance collapses quietly: no environment isolation, inconsistent security, unreadable audit trails, and reporting that turns into archaeology. Dataverse exists to stop that decay. It’s the Power Platform’s governed data backbone: relational schema with referential integrity, role/field-level security, managed environments (dev/test/prod), DLP/retention via Purview, reliable delegation, and first-class integration with Power Apps, Power Automate, Power BI, and Dynamics. The “Dataverse is expensive” objection is a cost illusion—SharePoint’s hidden rework, outages, and compliance risk cost more over time. Practical rule: use Lists for lightweight, short-lived, non-sensitive data; use Dataverse for anything relational, regulated, cross-app, or lon…
Guest: Mirko Peters
Listen to the Episode
Oct. 22, 2025

Your Fabric Lakehouse Is I/O Bound, Not CPU Bound – Here’s the Fix

Your Fabric and Power Platform workloads aren’t slow because of Spark or DAX—they’re slow because your data lives far from your compute. Managed storage adds network hops, caps IOPS, and taxes every read/write with latency. The fix isn’t “more nodes”; it’s proximity. Azure Container Storage v2 (ACStor v2) puts hot data on local NVMe inside your AKS nodes, exposing silicon-speed via a lean, CSI-based, ephemeral driver that stripes across all NVMe disks for obscene throughput (multi-GB/s, ~M IOPS). No SANs, no LVM, no etcd—just raw speed you’re already paying for in L-series/Dv6/NC VMs. Use it for Spark shuffles, Dataflows Gen2 staging, Direct Lake caches, and AI model weights; keep durable truth in Blob/managed stores. Result: 3–5× faster pipelines and dashboards, lower Fabric capacity burn, and dramatically cheaper I/O. ACStor v2 reframes storage: NVMe = racetrack (hot, fast, disposable), Blob = archive (cold, durable, cheap). Bring the bytes to the CPU and watch “overnight jobs” fini…
Guest: Mirko Peters
Listen to the Episode
Oct. 20, 2025

Dataverse Role Mistakes That Let Partners Download Your Customer List

Your Power App didn’t get “hacked”—it was over-permitted. Treating Dataverse like SharePoint (big buckets, broad roles) turns guest access into a data breach waiting to happen. Dataverse is a relational fortress built on granular privileges (Create/Read/Write/Delete/Append/Append To/Assign/Share), scoped access (User, Business Unit, Parent:Child, Organization), and Business Unit boundaries. One accidental Organization-level privilege on a guest or team role overwhelms every careful filter and exposes records across the environment.This episode shows the failure pattern (cloned roles, Parent:Child scope, team inheritance) and then the fix: isolate external users in their own Business Unit, build minimal guest roles from scratch, prefer Team ownership + Access Teams for precise sharing, apply Field-Level Security to sensitive columns, and automate join/leave via Entra ID. Close with governance: audit ownership and roles, enforce DLP with Purview, monitor high-scope changes, and run …
Guest: Mirko Peters
Listen to the Episode
Oct. 20, 2025

Before You Buy Power BI Premium, Fix This Data Model Problem

Your dashboards aren’t just slow—they’re expensive. Every bloated column, lazy import, and tangled relationship silently taxes your Power BI Premium capacity and your team’s time. That inefficiency adds up to real money—often five figures a year. The cure isn’t a plug-in; it’s architecture. Move from kitchen-junk-drawer models to a proper star schema: lean fact tables (events) surrounded by descriptive dimensions (product, customer, date). Keep relationships one-to-many, single-direction. Use surrogate keys, not “unique-ish” natural keys.Then impose DAX discipline: push transformations to Power Query (M) instead of calculated columns, favor columnar ops over row iterators, build clean base measures and layer logic with CALCULATE. Avoid bidirectional filters by default; reach for CROSSFILTER/TREATAS only when you truly mean it. Measure and tune with DAX Studio until refreshes finish in minutes, not hours.The payoff: lower capacity burn, faster refreshes, higher adoption, and re…
Guest: Mirko Peters
Listen to the Episode
Oct. 16, 2025

Stop Treating AI Like an App: How to Design It for GPUs and Data

AI is not “just another app” you park on general-purpose servers. Enterprise AI behaves like an ecosystem — volatile workloads, bursty data, exotic compute, and constant model evolution. That’s why so many AI pilots glow in the lab then die in production. The five tells that you’re no longer dealing with a normal workload: (1) you need horizontal scale, (2) accelerators like GPUs/TPUs matter, (3) data pipelines must flood continuously not trickle, (4) models mutate across versions and require versioning, observability and drift monitoring, (5) integration with legacy systems becomes the real bottleneck.The escape from “pilot / proof-of-concept death zone” is MLOps + orchestration — a Factory model — where DataOps, MLOps and GenAIOps operate from a unified command deck: templates, RBAC, private networking, GPU scheduling, AutoLake-style consistent data surfaces, and repeatability over artisanal hacking. The engine room is hardware + data + algorithms — and balance across those three …
Guest: Mirko Peters
Listen to the Episode
Oct. 11, 2025

Power BI Refresh Taking Hours? Switch to Direct Lake in OneLake

Microsoft Fabric fundamentally changes how Power BI handles data. With OneLake and Direct Lake, Power BI can now query lakehouse tables directly with performance similar to Import mode — without creating duplicate copies or maintaining complex refresh cycles.The winning Fabric pattern is simple:Dataflows Gen2 → Lakehouse → Pipelines → Semantic Model → Direct Lake reportOneLake becomes the governed vault (think: OneDrive for data). Purview delivers lineage + labeling from day one. Fabric admin controls let you enable it safely in trial capacity first — not in full production.Dataflows Gen2 hydrates the lakehouse. Pipelines keep it alive — and alert you when the 3am goblin breaks a step. Semantic models + Direct Lake then turn that hydrated lakehouse into fast, governed, analytics-ready Power BI.When you blend those moving parts — Fabric stops being a “feature” of Power BI and becomes the platform that powers it.
Guest: Mirko Peters
Listen to the Episode
Oct. 8, 2025

7 Sales Pipeline Mistakes That Make Good Leads Go Cold

In a recent podcast, Mirko Peters discussed the challenges of choosing between Synapse Link and Dataflow Gen 2 for Dataverse pipelines, emphasizing the importance of making informed decisions to avoid project failures. The conversation highlighted issues that arise when multiple teams create overlapping data links, leading to data duplication and governance chaos. Peters illustrated the differences between the tools: Synapse Link offers greater control and rollback capabilities but requires more complex setup, while Dataflow Gen 2 allows for quick, user-friendly data transformations but lacks robust governance features. He stressed the need for discipline in managing refresh schedules to prevent data loss and budget overruns. The discussion took place in the context of organizations needing to align their data management tools with their specific operational needs, advocating for a tailored approach rather than a one-size-fits-all solution. This topic is crucial as it addresses the po…
Guest: Mirko Peters
Listen to the Episode
Oct. 6, 2025

Stop Clicking 500 Times: Bulk Rename Power BI Measures with PBIP

In this episode, we dive into how to replace measures in Power BI using DAX, helping you simplify your data model, improve report performance, and create more accurate, maintainable calculations. We explain what measures are, why they’re essential for interactive Power BI reports, and how they differ from calculated columns. You'll learn when and why replacing a measure makes sense—whether for performance gains, model cleanup, or updating outdated logic.We walk through the step-by-step process of replacing measures in Power BI Desktop, including analyzing existing DAX, deciding between a new calculated column or a revised measure, and updating visuals to ensure accurate results. We also cover common troubleshooting issues like broken visuals, context-related errors, and performance bottlenecks—and how to avoid them.The episode also explores DAX fundamentals, including essential functions like CALCULATE, VAR, SUM, and SWITCH, along with best practices for writing clean, efficie…
Guest: Mirko Peters
Listen to the Episode
Oct. 5, 2025

How to Audit Power BI Semantic Models and Slash Capacity Spend

In this episode, we break down Microsoft Power BI’s pricing structure to help you understand the different license options, costs, and features available across the Power BI ecosystem. Whether you're comparing Power BI Free, Pro, Premium Per User, or Premium Per Capacity, this guide gives you a clear explanation of how each plan works and what it offers. We discuss how Power BI pricing supports everyone—from individual users exploring Power BI Desktop to large enterprises relying on dedicated capacity and advanced analytics.You’ll hear how each licensing model fits different business needs, what Power BI Pro includes for collaboration and report sharing, and when it makes sense to invest in Premium for scalability, AI-powered features, and improved performance. We also explore Power BI Embedded for app developers and explain how consumption-based pricing factors into capacity planning.The episode covers the key factors to consider when choosing a Power BI plan, including user …
Guest: Mirko Peters
Listen to the Episode
Oct. 4, 2025

Power BI vs Spreadsheets: When Dashboards Actually Win

In this episode, we explore how Power BI and Microsoft Teams work together to create a seamless, collaborative analytics experience. We break down how Power BI’s powerful data visualization capabilities integrate directly into Microsoft Teams, allowing organizations to share insights, discuss reports, and make data-driven decisions without switching apps.You’ll learn the basics of Power BI—connecting to data, creating reports, and publishing dashboards—alongside an overview of Microsoft Teams as a unified workspace for communication and collaboration. We explain how embedding Power BI reports into Teams channels enhances visibility, streamlines discussions, and keeps everyone aligned around the same analytics.The episode walks through practical steps for embedding, sharing, and accessing Power BI content in Teams using the Power BI app. We highlight how teams can collaborate in real time, discuss insights directly next to visualizations, and make reporting part of everyday wor…
Guest: Mirko Peters
Listen to the Episode