How to Design a Bronze Layer in Microsoft Fabric That Actually Scales
Aleksi Partanen - Master Microsoft Fabric
3,074 views
🔥 Learn Microsoft Fabric hands-on!
👉 https://www.skool.com/fabricforge
🔥 Ace Your Certification Exams – Free Practice Questions!
👉 https://certiace.com/
☕ Enjoying the Content? Support the Channel!
💖 https://buymeacoffee.com/aleksipartan...
👨💼 Follow Me on LinkedIn!
🔗 LinkedIn: aleksi-partanen
🚨 Join the Free Microsoft Fabric Community on Discord!
👉 Discord: discord
📚 Recommended courses on Datacamp:
Python Data Fundamentals: https://datacamp.pxf.io/c/6429510/312...
Data Analyst in Python: https://datacamp.pxf.io/c/6429510/312...
SQL Fundamentals: https://datacamp.pxf.io/c/6429510/312...
Power BI Fundamentals: https://datacamp.pxf.io/c/6429510/291...
PL-300: Microsoft Power BI Data Analyst: https://datacamp.pxf.io/c/6429510/312...
Data analyst in Power BI: https://datacamp.pxf.io/c/6429510/312...
More courses: https://datacamp.pxf.io/c/6429510/207...
These are affiliate links. If you sign up through them, I may earn a small commission at no extra cost to you. Thanks for supporting the channel.
📂 Access Learning Materials (Files, Code, etc.)
💾 Download helpful resources here:
🔗 https://drive.google.com/drive/folder...
🎓 More of My Content:
🎥 DP-700 Microsoft Certified: Fabric Data En...
🎥 Microsoft Fabric Data Engineering
🎥 Microsoft Fabric Tutorials
🎥 DP-600 Microsoft Certified: Fabric Analyti...
🎥 Learn Azure Data Factory in 2026 - Full Co...
🎥 Learn Microsoft Fabric Data Pipelines in 2...
🎥 Learn Microsoft Fabric Notebooks in 2026 -...
🔗 All the Videos:
@aleksipartanentech
👑 Video Guests 👑
Teemu Multanen
LinkedIn: teemu-multanen
@teemumultanendata
In this video, Teemu Multanen and I break down how to design a bronze layer in Microsoft Fabric that actually scales. We discuss why bronze is the most important medallion architecture layer to get right, whether a lakehouse or a warehouse is the better bronze data store, and whether you should have one shared bronze lakehouse or one per source system.
Then we walk through an example metadata-driven framework in Fabric: landing data as raw files in a timestamped folder structure, loading it to Delta tables with a schema per source system, tracking change history with an SCD2 style load, and driving the whole thing with JSON configuration files and reusable pipelines. Along the way we share real project experiences with messy legacy source systems, business keys, and where these designs start to break at enterprise scale.
Topics covered:
Why the bronze layer is the most important layer to get right
Lakehouse vs warehouse as the bronze data store
One shared bronze lakehouse vs one per source system (discoverability, blast radius, the 1,000 item workspace limit)
Landing data as raw files: formats and a timestamped folder structure
Archiving raw extracts so you can rebuild downstream layers
Delta tables organized with a schema per source system
SCD2 / change history loads in bronze and when to use them
Primary keys vs business keys in messy legacy source systems
Metadata-driven pipelines: Lookup + ForEach over a configuration file
JSON configuration files vs configuration tables
Extending configurations toward data contracts and data products
One generic load notebook vs per-source notebooks
#microsoft #microsoftfabric #deltalake #lakehouse #dataengineering
13 Comments
👉 https://certiace.com/
How to Design a Bronze Layer in Microsoft Fabric That Actually Scales
Aleksi Partanen - Master Microsoft Fabric
3,074 views
🔥 Learn Microsoft Fabric hands-on!
👉 https://www.skool.com/fabricforge
🔥 Ace Your Certification Exams – Free Practice Questions!
👉 https://certiace.com/
☕ Enjoying the Content? Support the Channel!
💖 https://buymeacoffee.com/aleksipartan...
👨💼 Follow Me on LinkedIn!
🔗 LinkedIn: aleksi-partanen
🚨 Join the Free Microsoft Fabric Community on Discord!
👉 Discord: discord
📚 Recommended courses on Datacamp:
Python Data Fundamentals: https://datacamp.pxf.io/c/6429510/312...
Data Analyst in Python: https://datacamp.pxf.io/c/6429510/312...
SQL Fundamentals: https://datacamp.pxf.io/c/6429510/312...
Power BI Fundamentals: https://datacamp.pxf.io/c/6429510/291...
PL-300: Microsoft Power BI Data Analyst: https://datacamp.pxf.io/c/6429510/312...
Data analyst in Power BI: https://datacamp.pxf.io/c/6429510/312...
More courses: https://datacamp.pxf.io/c/6429510/207...
These are affiliate links. If you sign up through them, I may earn a small commission at no extra cost to you. Thanks for supporting the channel.
📂 Access Learning Materials (Files, Code, etc.)
💾 Download helpful resources here:
🔗 https://drive.google.com/drive/folder...
🎓 More of My Content:
🎥 DP-700 Microsoft Certified: Fabric Data En...
🎥 Microsoft Fabric Data Engineering
🎥 Microsoft Fabric Tutorials
🎥 DP-600 Microsoft Certified: Fabric Analyti...
🎥 Learn Azure Data Factory in 2026 - Full Co...
🎥 Learn Microsoft Fabric Data Pipelines in 2...
🎥 Learn Microsoft Fabric Notebooks in 2026 -...
🔗 All the Videos:
@aleksipartanentech
👑 Video Guests 👑
Teemu Multanen
LinkedIn: teemu-multanen
@teemumultanendata
In this video, Teemu Multanen and I break down how to design a bronze layer in Microsoft Fabric that actually scales. We discuss why bronze is the most important medallion architecture layer to get right, whether a lakehouse or a warehouse is the better bronze data store, and whether you should have one shared bronze lakehouse or one per source system.
Then we walk through an example metadata-driven framework in Fabric: landing data as raw files in a timestamped folder structure, loading it to Delta tables with a schema per source system, tracking change history with an SCD2 style load, and driving the whole thing with JSON configuration files and reusable pipelines. Along the way we share real project experiences with messy legacy source systems, business keys, and where these designs start to break at enterprise scale.
Topics covered:
Why the bronze layer is the most important layer to get right
Lakehouse vs warehouse as the bronze data store
One shared bronze lakehouse vs one per source system (discoverability, blast radius, the 1,000 item workspace limit)
Landing data as raw files: formats and a timestamped folder structure
Archiving raw extracts so you can rebuild downstream layers
Delta tables organized with a schema per source system
SCD2 / change history loads in bronze and when to use them
Primary keys vs business keys in messy legacy source systems
Metadata-driven pipelines: Lookup + ForEach over a configuration file
JSON configuration files vs configuration tables
Extending configurations toward data contracts and data products
One generic load notebook vs per-source notebooks
#microsoft #microsoftfabric #deltalake #lakehouse #dataengineering
👉 https://www.skool.com/fabricforge