How a side project became a go-to resource, and what the analytics taught us about how people actually learn.
There is a moment every builder dreads.
You spend months on something. You write the lessons. You design the flow. You test every example on Databricks Free Edition so nobody has to pay a cent. Then you hit publish.
And you wait.
For BricksNotes, the wait did not last long.
In the last 90 days, BricksNotes saw 8,000 pageviews, 2,700 unique visitors, and 3,100 sessions. People stayed for an average of 6 minutes. The bounce rate was 45.7%, which for a technical site is genuinely good. Pages per session sat at 2.6.
But the number that matters most is bigger than any of those.
Over 40,000 data engineers across the world are now using BricksNotes and its resources to learn and apply data engineering with Databricks.
That is not a vanity number. It is people opening notebooks, writing PySpark, debugging pipelines, and building real skills.
Tutorials teach syntax. BricksNotes was built to teach thinking.
There are hundreds of Databricks tutorials online. YouTube is full of them. Udemy has courses. Databricks itself publishes excellent documentation.
So why did 40,000 people show up here?
Because most resources teach syntax. They do not teach thinking.
You can watch a 45-minute video on Spark DataFrames and memorize df.select().filter().groupBy(). But the moment someone asks you to design an incremental pipeline with SCD Type 2 handling on a partitioned Delta table, you freeze. You learned the words but not the logic.
The gap between finishing a tutorial and building production pipelines is huge. Almost nobody is bridging it.
That is what BricksNotes was built to do.
Every lesson starts with doing, not reading.
The idea is simple. You learn data engineering by engineering data. Not by watching someone else. Not by memorizing syntax cards. Not by passing a multiple-choice test.
You open a Databricks notebook. You write code. You break something. You figure out why. You fix it. That is the loop.
And every lesson runs on Databricks Free Edition. No cloud bill. No enterprise license. No credit card. Just open your browser and start.
This was a deliberate choice. The biggest barrier to learning Databricks is not complexity. It is access. When someone can go from zero to their first Spark job in under an hour, with no financial risk, everything changes.
BricksNotes covers what a working data engineer actually needs:
Each topic builds on the last. There is a reason Delta Lake comes after DataFrames. There is a reason Unity Catalog comes near the end. The order matters because real understanding is cumulative.
The dashboard told a story we did not expect.
The busiest day was Friday. Not Monday. Not Wednesday. Friday. Data engineers are spending Friday afternoons leveling up. That says something about this community.
Visitors stay 6 minutes on average. On a technical site, that is a lot. It means people are reading, not skimming.
The engagement rate is 54%. More than half of visitors interact with the content in a meaningful way.
2.6 pages per session tells us the curriculum is working. People read one lesson, then open the next, then the next.
Most visitors come from the US, but the reach is global. Quality free resources for Databricks are needed everywhere.
The Databricks ecosystem is moving fast. Unity Catalog is the production-ready standard for governance and Iceberg cataloging. Delta Lake is the default storage layer. New tools like the Databricks connector for Microsoft Power Platform are shipping native Databricks connectors so business teams can build apps on warehouse data without writing SQL.
Every advance like this makes the data layer more important, not less.
When business teams can self-serve dashboards on top of your tables, the quality of those tables becomes the bottleneck. Poor schema design, missing partitioning strategies, no quality checks, undocumented transformations. These problems do not go away with better tooling. They get amplified.
Engineers who understand Delta Lake deeply, who design schemas that evolve gracefully, who use Unity Catalog properly, who build Lakeflow Declarative Pipelines (formerly Delta Live Tables), those engineers become hard to replace.
That is what BricksNotes teaches. Not just the how. The why.
The 40,000 number is just a beginning.
More advanced lessons are coming, on the topics readers keep asking for. Complex streaming patterns. Optimization techniques. Real-world pipeline case studies. We are also tracking the evolution of Delta Sharing into OpenSharing for cross-platform asset exchange.
The cheat sheet library keeps growing. The PySpark and SQL cheat sheets are already among the most visited pages.
Community-driven content based on the questions real data engineers face every day.
And the fundamentals will stay free. The basics of data engineering should not sit behind a paywall. The industry needs more skilled data engineers, and the fastest way there is to remove every barrier.
If you are learning Databricks, or thinking about it, or stuck somewhere in the middle, BricksNotes was built for you.
The engineers getting hired and promoted right now are not the ones with the most certificates on the wall. They are the ones who can design a pipeline from scratch, explain every decision, and debug it when something breaks at 2 AM.
That is what we teach here. Come build with us.