See how it works in five minutes.

01

Pick a tour

One tour per studio, each on the studio's first scenario, replayed from real work rather than a mock-up.

1. 01Studio 01 · Data PipelineDaily orders ETLBatch or streaming, what the quality gate does, how the sink writes — then the checks read every setting on your DAG.Draw the DAGConfigure every nodeGet it reviewedTake the tour →
2. 02Studio 02 · Data ModellingMarketplace core entitiesWhere an order’s lines live, what stops a double charge, what an order remembers — then your ERD is built, seeded and queried.Set grain and keysSeed and query itDefend the designTake the tour →
3. 03Studio 03 · Spark PlaygroundTwo counts, two scansCall which cells start a Spark job, run them, then read the job DAG, the plan and the task times Spark recorded.Predict the jobsRead the DAG and planExplain the shuffleTake the tour →

02

The same loop, every studio

The way a senior engineer works through a design review, and the way an interviewer walks you through one.

1. 01DecideMake the real calls: the grain, the write mode, which line starts a Spark job.
2. 02CheckSee what the review or Spark itself says about your call, with the evidence.
3. 03DefendExplain it out loud, the way the follow-up question will ask you to.

03

Or skip the tour, start building