Learning Paths by Role¶
Three guided routes through the handbook, one for each role: analytics engineer, data platform engineer and AI data engineer.
The guides are organised by topic. These paths reorder them by what a role does day to day, and pair each stage with a hands-on lab and a way to check that you are ready to move on. Every lab uses the same e-commerce dataset, so the work carries over from one stage to the next.
| Analytics engineer | Data platform engineer | AI data engineer | |
|---|---|---|---|
| You own | Trusted, modelled data that the business queries | The systems that move, store and run data reliably | The data and evaluation behind LLM applications |
| Main tools | SQL, a warehouse, dbt, BI | Kafka, Spark, Airflow, Kubernetes, Terraform | Embeddings, vector search, RAG, evals, agents |
| Start here if | You write SQL and want to model and test it well | You like infrastructure, streaming and operations | You build or feed LLM applications with private data |
| Labs | 01, 02, 08, 06 | 10, 04, 05, 03, 09, 06 | 07, 08 |
| Lab time | About 5 to 7 hours | About 9 to 12 hours | About 2 to 3 hours |
How to use a path¶
- Read the guide, then do the lab. The guides explain the ideas and show the patterns. The labs make you build and debug them, and each lab checks its own results.
- Use the checkpoint at the end of each stage. If you cannot answer a question without looking, go back to that guide's Common Pitfalls and Interview Questions sections.
- Skip what you already know. Every stage lists what it covers, so you can move past a stage you have done at work.
- Finish with a capstone. The capstone projects combine several stages into one pipeline.
If none of these fits, the topic-based paths on the home page cover beginners, warehouses, Spark, streaming and AI engineering. The interview roadmap maps every topic to the guides that cover it.
Missing a role? Open a topic request.