Microsoft Fabric Architecture Podcast

Matthias Falland

Architecture decisions for Microsoft Fabric. Anonymized real customer scenarios, cost realism, counter-arguments included. Weekly episodes aligned with Fabric Friday recordings.

Play

Open in

20 episodes
Avg 9 min
English

Episodes 20

Newest Oldest Longest

All20 Unplayed20 In progress0 Played0

#36
September 4 · 10 min
The Folder That Carries Four Boundaries — Fabric Workspaces
A Fabric workspace isn't a folder — it's a cut through OneLake carrying every access rule, capacity bill, and residency commitment. Fabia and Matthias trace the boundaries hiding inside one object, starting with the documented fact that Contributor overri
- Transcript
- Chapters
#35
August 28 · 9 min
When Your Graph Query Lies to You — Fabric Graph
Graph in Fabric materialises a snapshot of your data's relationships the moment you save the model. That one decision explains the stale queryset, the failed-to-load error and the corruption right after a save. The 64 MB truncation is a separate limit, an
- Transcript
- Chapters
#34
August 21 · 10 min
The Radio Button That Sets Your Security Architecture — API for GraphQL
Fabric generates a GraphQL API from your warehouse in under a minute. But the connectivity dialog you see once decides your security posture, your RLS enforcement, and whether your production deployment silently reads development data. We trace the conseq
- Transcript
- Chapters
#33
August 14 · 9 min
The Permission Broker That Outlives You — Org Apps in Fabric
Org apps solved the decade-old broken-visual problem by managing permissions along the dependency chain. But the same engine that grants access on your behalf can leave records behind that you no longer have the rights to remove, and a single audience tog
- Transcript
- Chapters
#32
August 7 · 9 min
When Publish Succeeds and Nothing Changes — Power BI Apps
Power BI workspace apps promise a clean publish boundary between builders and consumers. But the mechanism that makes that boundary work — the copy, the separate IDs, the audience arithmetic — carries five traps that all fail in silence.
- Transcript
- Chapters
#31
July 31 · 10 min
The Database Hiding Inside Your Scorecard — Fabric Scorecards
A scorecard silently creates its own semantic model the moment you hit New. That one architectural fact — that it's a stateful writer, not a read-only lens — explains why RLS doesn't carry over, why refresh breaks upstream, and why rollups leak across per
- Transcript
- Chapters
#30
July 24 · 10 min
One Expression Halves Your Data Ceiling — Paginated Reports
Paginated reports carry no data model, and that one design decision explains every limit teams hit in production. This episode traces the consequences from silent data drops to a tenant-wide security door that only opens one way, and lands on the question
- Transcript
- Chapters
#29
July 17 · 9 min
The Dashboard That Can't Filter — Power BI Dashboards
A Power BI dashboard can show tiles from six different semantic models. It cannot filter a single one of them. We explain why the architecture makes that tradeoff inevitable, and what it means for choosing between dashboards and reports.
- Transcript
- Chapters
#28
July 10 · 10 min
The Bucket Nobody Reads — Power BI Reports
Every visual on a report page fires a DAX query. When twenty visuals queue behind a parallelism cap, the bottleneck hides in Performance Analyzer's least-read column — and the escape hatch everyone reaches for can silently produce wrong numbers.
- Transcript
- Chapters
#27
July 3 · 9 min
When Direct Lake Goes Quiet — Power BI Semantic Models
Direct Lake promises no refresh and VertiPaq speed at lake scale. But its failure modes are silent — it degrades instead of breaking. We pull apart framing, transcoding, and the fallback tax nobody talks about.
- Transcript
- Chapters
#26
June 26 · 9 min
The Column Named C1 — Fabric Data Agent
A Fabric Data Agent generates queries from your column names, not your intentions. This episode pulls apart the grounding pipeline to show where accuracy lives and dies — and why the schema investment you skipped three years ago just became urgent.
- Transcript
- Chapters
#25
June 19 · 9 min
The Function That Bills by the Row — Fabric AI Functions
AI Functions let you call GPT from a SELECT statement with no API key. Zero infrastructure — until you check the capacity meter. Two billing streams, no result cache, and a default model eleven days from retirement.
- Transcript
- Chapters
#24
June 12 · 9 min
The AI That Doesn't Send an Invoice — Copilot in Fabric
Copilot in Fabric has no per-user license fee. But every request draws from the same capacity pool as your pipelines and reports. We dig into what 'included' actually costs — and when it throttles production.
- Transcript
- Chapters
#23
June 5 · 8 min
When Autolog Stops Logging — Fabric Experiments
Fabric's MLflow autologging promises zero-effort experiment tracking — until you train with LightGBM or XGBoost and discover your metrics column is blank. We pull apart the framework gap, the compounding exclusive flag, and the Git backup that isn't one.
- Transcript
- Chapters
#22
May 29 · 8 min
Your Endpoint Falls Asleep After Five Minutes — Fabric ML Models
Fabric ML Model wraps an MLflow registry inside your workspace — batch scoring to Power BI with zero refresh sounds ideal. But the real-time endpoints are Preview, Git tracks metadata only, and there's no GPU. When does this actually replace Azure ML, and
- Transcript
- Chapters
#21
May 22 · 10 min
The Contract That Drops Your Data — Fabric Schema Registry
Fabric's Schema Registry promises type-safe streaming pipelines. But non-conforming events vanish without a dead-letter queue, compatibility checks don't exist, and you can't retrofit schema mode on a running Eventstream. We dig into what the contract act
- Transcript
- Chapters
#20
May 15 · 9 min
The Stop Button That Doesn't Stop — Data Activator
Fabric Activator's no-code alerting looks clean on the surface. But the entity model underneath hides a billing mechanic that catches every team once: a Stop button that doesn't stop the meter. We pull it apart.
- Transcript
- Chapters
#19
May 8 · 10 min
The Alert That Saved Successfully — Real-Time Dashboard
Real-Time Dashboard puts an alert button on every tile and confirms the save every time. We trace how the Kusto-native monitoring surface actually queries, how its permission model quietly separates viewing from querying, and where the alerting pipeline s
- Transcript
- Chapters
#18
May 1 · 9 min
The Catalog Nobody Asked For — Real-Time Hub
Real-Time Hub puts the same twenty-five connectors in a second UI nobody requested. We dig into why this auto-provisioned catalog becomes load-bearing the moment streams outlive their builders — and the Activator cost trap hiding behind every alert.
- Transcript
- Chapters
#17
April 24 · 10 min
The KQL Function Call Nobody Writes — KQL Queryset
Most teams learn KQL through the explain keyword, which translates their SQL into syntactically correct KQL — and stops at the subset SQL can express. We trace the structural gap between bin-plus-summarize and make-series, and why the anomaly detection mo
- Transcript
- Chapters

Showing 1–20 of 20 episodes