Nikola Ilic

Nikola Ilic

@datamozart.bsky.social

I'm making music from the data🎶 #PowerBI addict, Microsoft Data Platform MVP, O'Reilly Instructor, Pluralsight Author, blogger, speaker, learner...Father of 2, Barca and Messi fan⚽

1.2K followers254 following631 posts

Pinned

A moment of disbelief! A little more than a year ago, when I started writing a book about #MicrosoftFabric, I couldn't have imagined that the day would come when I would have a physical copy in my hands...

Inspired by Josue “Josh” Bogran's animated infographic about the compute options in Databricks, I've decided to create one for #MicrosoftFabric. 5 engines -> 5 different workloads -> 5 different sets of use cases!

Thank God I waited, since there were many cool new #MicrosoftFabric features in the meantime to support the article's plot (who said: IDENTITY columns☺️) As a bonus point, it also includes my thoughts about where SCD Type 2 should sit in the medallion design pattern.

NEW BLOG POST! Recently, I went down this rabbit hole twice within a week:) If you’ve tried to check how much storage your #MicrosoftFabric warehouse tables consume, you’ve probably discovered that the usual SQL Server approaches don’t work. data-mozart.com/how-to-check...

data-mozart.comHow to check table sizes in a Microsoft Fabric warehouse - Data MozartIf you've tried to check how much storage your Fabric warehouse tables consume, you've probably discovered that the usual SQL Server approaches don't work. This post explains why, what else doesn't wo...

NEW ARTICLE! The medallion design pattern is probably the most widely adopted pattern of the lakehouse era, and in almost every implementation I've seen, all three layers are physically materialized. But why exactly is the silver layer a set of tables? data-mozart.com/dont-stop-ma...

data-mozart.com(Don't) Stop materializing your silver layer! - Data MozartWhy is your silver layer a set of tables? Challenging the fully materialized medallion pattern - with benchmark numbers and reproducible demo scripts.

The most expensive code your data team ships this year won't be written by a human. It'll be generated by AI assistant. It'll look correct, but it'll quitely drop 14% of your rows because the join type was wrong. AI-assisted data engineering is here. courses.data-mozart.com

courses.data-mozart.comAI-Native Data Engineering with Microsoft FabricA live 4-hour, code-first workshop on Fabric implementation, agentic scaffolding, metadata-driven ingestion, validation, cost, and production trust.

Humbled and honored to be recognized as Microsoft Data Platform MVP for the 6th consecutive year😊 Especially honored to receive the award in the Real-time Intelligence category for the first time! #MVPbuzz #MicrosoftFabric #CommunityRocks

NEW BLOG POST! “AI will replace data engineers”… “Agents write better code than humans”… “If you’re not vibe coding, you’re already behind”… “Nikola, should we let AI write our pipelines?” Yes… but not the way you probably think. data-mozart.com/liar-liar-wh...

data-mozart.comLiar, liar! What to do when AI writes your pipelines confidently wrong? - Data MozartShould we let AI write our pipelines? YES...But not the way you probably think.

NEW BLOG POST! Starting Aug 1st, Microsoft is changing how your Fabric Warehouse (and SQL EP of the lakehouse) workloads are metered within the capacity. I cover what's changing (and what stays the same) and how the new model compares to Snowflake and Databricks. data-mozart.com/one-minute-t...

data-mozart.comOne-minute trap! How new Fabric Warehouse CU metering affect your workloads - Data MozartMicrosoft is moving Fabric Warehouse CU metering from per-query CPU time to per-workspace virtual-node time. That brings Fabric closer to Snowflake and Databricks, but it also changes the cost equatio...

Nikola Ilic reposted

This Thursday, I'll be talking at the Microsoft Fabric UK User Group, about transitioning from data analyst to data engineer and upgrading your skills from Dataflows Gen2 to notebooks. It's free and it's online, so RSVP: www.meetup.com/microsoft-po...

meetup.comTransition from Data Analyst to Data Engineer, Thu, Jun 25, 2026, 6:00 PM | MeetupJoin our Microsoft Fabric - UK User Group for our latest webinar. This session we are excited to announce our host Microsoft MVP Leon Gordon and co-host Microsoft MVP Prag

NEW SUBSTACK ARTICLE! If I had a euro every time someone asked me “should I use Warehouse or Lakehouse in #MicrosoftFabric?” - I’d have enough to fund an F64 for a month:) And, my answer has been the same for a while now: you’re asking the wrong question! datamozart.substack.com/p/same-lake-...

datamozart.substack.comSame lake, different shores: Warehouse vs. Lakehouse in 2026If I had a euro for every time someone asked me "should I use the Warehouse or the Lakehouse?" - I'd have enough to fund a month of exclusive vacation...

I've just published a reference framework covering the full stack: capacities, domains, workspaces, folders, items, schemas, and tables. The opinionated patterns I'd use, and the reasoning behind each one.

I recently delivered my "Chief Data Alchemist - Crafting Medallion the right way!" session at SQL Saturday Atlanta, where I used publicly available NBA data to showcase how data flows from raw JSON to fine, gold aggregated tables. I was using notebooks for all the transformations

Now, with Materialized Lake Views going GA, I decided to rebuild the same flow using MLVs. Some findings will definitely surprise you! Stay tuned, the video is coming soon:) What do you say about Claude's work? I like the style!

NEW SUBSTACK ARTICLE! If you ever tried to debug a silver layer that didn't update because a notebook failed 3h ago, you know why materialized lake views matter. The promise is simple: what if your entire medallion pipeline could be a few SELECT statements? datamozart.substack.com/p/materializ...

datamozart.substack.comMaterialized Lake Views in Microsoft Fabric: When your medallion fits in a SELECT statementFive surfaces collapsed into one declarative layer. Here's the full story of Materialized Lake Views in Microsoft Fabric - from syntax to the new GA capabilities announced at FabCon Atlanta

I spent the last few days writing a full deep dive on them, covering: 1️⃣ What MLVs actually are (and what they replace) 2️⃣ The Spark SQL syntax + the new PySpark @fmlv decorator API 3️⃣ How optimal refresh decides between skip, incremental, and full

4️⃣ The 5 changes that moved MLVs from preview to production-ready at FabCon Atlanta 5️⃣ When MLVs shine and when to reach for notebooks instead 6️⃣ The fine print: limitations you should know before committing #MicrosoftFabric #DataEngineering #Lakehouse

Ok, the cat is out of the box - full-day tutorials for #FabCon Europe in Barcelona have been announced, and I can't be more excited to join the amazing speaker line-up! I'll be teaching how to build a production-ready lakehouse in #MicrosoftFabric.

View full thread

What you'll do instead: build a complete Microsoft Fabric solution end-to-end, with your hands on the keyboard the entire day. Real ingestion patterns. Real transformations. Real trade-offs you'll face on Monday morning back at work.

Ok, the cat is out of the box - full-day tutorials for #FabCon Europe in Barcelona have been announced, and I can't be more excited to join the amazing speaker line-up! I'll be teaching how to build a production-ready lakehouse in #MicrosoftFabric.

I've created this workshop because I was bothered:). Here's what bothers me about most data engineering workshops: You show up excited, spend the day watching slides, nod along to concepts you already half-knew, and leave with a bunch of papers full of diagrams you'll never open again.

All ready and set for my #MicrosoftFabric Data Engineer Bootcamp (DP-700) live training at @oreilly.bsky.social! Skills measured for the exam will be updated next Monday, but you are already covered, as the Bootcamp includes the fresh curriculum and all the topics. See you soon!

NEW SUBSTACK ARTICLE! You’ve probably spent last year hearing about Lakehouses and Warehouses until your ears are ringing. If a Lakehouse is a massive library and a Warehouse is a high-end bank vault, the Eventhouse is a 24-hour Air Traffic Control Tower. datamozart.substack.com/p/the-new-ho...

datamozart.substack.comThe new house on the block: Why you need a smoke alarm, not just a newspaperYou’ve probably spent the last year hearing about Lakehouses and Warehouses until your ears are ringing. This article helps you understand in which scenarios you might need an eventhouse