Figuring Out Fabric: Learn Fabric in 30 minutes.

T. from Data Rocks talks about how data viz is a tiny subset of information design. The key is to focus less on just charts, but more about how the data is communicated and received. We talk about how what the user does with it separates a pile of charts from a successful design flow. I found this conversation helpful to understand it means to be good at data viz.

Links

• Stop Solutioneering
• Financial Intelligence for IT People
• Solving the Right Problems by Kurt Buhler
• Don't Make Me Think
• The Pepsi Paradox
• www.datarocks.co.nz

In this episode, Steph Locke covers a wild career from data science consultant to startup owner to Microsoft manager. We talk about what’s required to work in data science. We also talk about the interaction of large language models and coding. Finally, we talk about adjusting to Power BI and Fabric.

Links

• Statistical t-test
• Statistical r-correlation
• Base rate fallacy
• ROC (statistical test)
• Roc (mythical bird)
• Why your AI abstract was rejected
• Against Power BI dogma

In this episode with Ellen Burns-Johnson, we talk about when we worked on a cloud game for Azure Synapse, ADF, Power BI. A big lesson learned from that is a big challenge today for Fabric is getting different teams and personas to communicate with each other; it's not just about the technology. Ellen's first impression is "it seems like Microsoft is trying to 'product away' the communication layer". 

Links

• Scott Hanselman and Abstractions
• Henry Forde never said anything about faster horses
• Big data is dead
• January 2025 update on Reddit
• Mayor uses deeply disturbing AI image
• Power BI, AI, and Ethics
• Deepseek thinks it is ChatGPT
• Kind and wicked learning environments
• Body Doubling
• Parallel Play

In this Episode Els Van Vessem talks about the challenges of implementing Fabric in the government. In it they talk about doing proof of concepts with Fabric and the limitations when your data is confidential, protected and on-premises. Data sovereignty often means sticking to a hybrid approach. We also discuss the challenges of similarly named products with slight differences.

Links

• Veteran's affairs hack
• Police force hack
• Fabric Mirroring
• Fabric Warehouse T-sql surface area
• Freedom of Information Act
  
  

Apologies for the audio; I stupidly forgot to double check my mic this episode.

In this episode, Erin Dempster gives us an outside view of fabric focused on CI/CD. We talk about both deployment pipelines and devops pipelines and how she uses both tools in concert. This episode is interesting because it touches on the challenges of integrating a variety of data sources for an insurance company and using CI/Cd to keep everything in sync.

• PBIP file format
• PBIR file format
• Deployment Pipelines
• Devops Pipelines
• Automating tests for broken visuals

In this episode, Stephanie Bruno talks about semantic link and semantic link labs, which allow you to better manage your Power BI resources with Fabric notebooks. Semantic allows you to query and work directly with your semantic model. Semantic Link Labs allows you to automate running the best practices analyzer against your model. I've heard nothing but great things about it.

Links

• Semantic Link
• Semantic Link Labs
• Fabric.guru by Sandeep Pawar
• Fabric Notebook Utils
• Semantic best practices
• Report Analyzer
• SNL Shimmer skit
• ChatGPT to document measures

In this episode, Krystina Mishra talks about being an accidental Fabric admin. She talks about the challenges of being part of a centralized IT team that operational and business teams. She talks about the challenges of how everything with Dynamics 365 is slightly different than every other data source and how everything is convoluted with Synapse. She also talks about how there are nuanced differences between a P1 and an F64.

Links

• Dynamics Great Plains schema
• How to Have Confidence and Power in Dealing with People
• What is a thermocline
• Fabric surge protection
• Fabric capacity metrics app
• Satisfactory video game
• What is Brownian Motion
• Fundamentals of Microsoft Fabric
• Kansas City SQL User Group

In this Episode, Frank Geisler explains Realtime Intelligence in Microsoft Fabric. We learn how RTI is its own thing in Fabric and is not directly backed by Parquet like a Lakehouse is. We also dig into the distinction between real-time analytics and real-time intelligence. The latter is not just reporting but being able to trigger activity based on it.

Links

• Fabric Realtime Intelligence
• KQL language
• Kusto Detective
• Realtime Dashboard
• Data Activator
• Frank's Tutorials
• Microsoft Sentinel and KQL

In this episode, Ginger Grant talks about the differences between warehouse and lakehouses in Fabric. We talk about how Warehouses make more sense if you are doing a lot of ad-hoc querying. In most other cases, Lakehouse will be easier and fewer steps.

Links

• Lakehouse versus Warehouse decision guide
• Delta lake vacuum
• Polaris engine behind Fabric Warehouse
• Lakehouse versus Warehouse (Advancing Analytics)
• Fabric surge protection (preview)

In this episode, Kellyn Gorman talks about the challenges of getting data out of legacy systems (i.e. relational data bases) into Fabric. She explains that whoever hosts the data wins. She talks about often content talks about the golden path or focuses on the marketing content, but it's much rarer to see content that deals with the difficult edge cases.

We talk about how despite being Software as a Service, in order to learn Fabric you will need to learn networking, authentication, and infrastructure more broadly. We talk about how it's impossible to completely get away from hardware and infrastructure.

• Ballmer does an about face on Linux
• Lamp stack
• James Serra on ways to load data into Fabric
• Eugene's benchmarking post, inspired by Jame's post
• AzCopy tool
• Scott Hanselman on abstractions
• Wolf fence algorithm of troubleshooting

Cathrine explains how there isn't a single solution for architecting your data lake with Microsoft Fabric. We walk through all the different moving pieces of getting started with Fabric and lakehouses. Catherine touches on some different ways of implementing medallion in Fabric. She also makes the point Medallion is not the same as Dev / QA / Prod. Lastly, we talk about source control and branching workspaces in Fabric.

Links

• Fabric February
• Medallion architecture (Microsoft)
• Medallion Architecture (Databricks)
• Fabric concepts (capacity, workspaces, licensing)
• Ouroboros

In this Episode, I interview Kristyna Ferris about the different types of data Movement in Microsoft Fabric. Specifically, we talk about gen 2 dataflows, data pipelines, and Spark notebooks. We see how you start simple and work your way up. Kristyna shares the "faucets first" approach at P3 adaptive.

Links

• Yak shaving
• Rapid development
• Datatypes in Fabric Data Warehouse
• VARCAHR(MAX) preview
• Microsoft decision guide
• Semantic Link - Sempy
• Power Bi file import benchmarks
• Semantic Link Labs
• Data on Wheels blog