Same weekly rhythm, same mentor, same honest approach — applied to two different destinations. Start with whichever matches where you want to end up.
Predictable and repeatable, so the only thing that changes week to week is the topic — never the routine. Both bootcamps below follow this exact structure.
A 1–1.5 hour live session to kick off the week's topic — lecture and demo-led, no coding pressure in the room.
Work through the material at your own pace during the week, on your own schedule.
A take-home assignment tied directly to what you just learned, building on the same dataset every time.
The core discipline behind every data-driven system. Fifteen weeks, one running dataset, and the exact tools working data engineers use every day — taught by someone who runs the real thing.
Module 1 starts at the far left. By the capstone, you're building the entire chain yourself.
Where every real dataset starts.
Getting data in from an app DB, an API, a log file.
Putting it somewhere sensible — a warehouse.
Cleaning and reshaping it into something usable.
Running it on a schedule, reliably, with alerts.
What everyone downstream is actually waiting for.
And the same ones you'll use in your capstone project.
Housekeeping, who's in the room, and what the next 15 weeks actually look like.
The core language skills every later session builds on.
How data is stored and queried, and why Postgres is the foundation for the rest of the program.
Writing real queries and structuring data the way production systems expect.
The concepts that separate a script from a pipeline: reliability, idempotency, and design.
Scheduling and monitoring ingestion so it runs reliably — before you touch transformation.
Transforming raw tables into clean, tested, documented models.
Build the entire chain yourself, end to end, on the dataset you've used all along.
The data discipline behind real AI systems. Fifteen weeks focused on what actually makes AI applications work — clean sources, chunking, embeddings, and retrieval — not just calling a model and hoping for the best.
Garbage in, garbage out still applies — it just looks different. Module 1 starts at the far left. By the capstone, you're building a working, grounded AI system yourself.
PDFs, wikis, tickets — messy and unstructured.
Splitting content into retrievable, sized units.
Turning text into vectors a model can search.
Storing and indexing embeddings at scale.
Pulling back the right context, not just any context.
An AI answer the business can actually trust.
Including one you'll already know if you've taken Data Engineering Foundations first.
Who's in the room, and why AI engineering still starts with data, not a model.
Calling models, structured outputs, and the basic patterns every later session builds on.
How text becomes vectors, and why that choice quietly decides your retrieval quality.
Storing, indexing, and querying embeddings — in Postgres, not a new system to learn.
Chunking strategy, retrieval, and grounding model responses in your actual data.
Building the datasets that tell you whether your AI system is actually working — not just demoing well.
Scheduling embedding jobs, refreshing indexes, and monitoring pipeline health in production.
Build a working, evaluable RAG system end-to-end, on a dataset of your choice.
Tell us which program fits and where you're starting from — we'll tell you exactly what's involved.