AI Engineering: LLMs, agents & RAG, done properly
Driving and building with AI is now core engineering. This is the dedicated chapter most courses skip — LLM foundations, prompt engineering, agents, RAG, evaluations and the tooling companies actually run.
Because “use an LLM” isn't a skill plan
Companies need engineers who can build reliable AI features — not just call an API. So we teach it deliberately, and make you prove it.
The chapter courses skip
Most programs teach a framework and stop. We dedicate a full track to LLMs, agents, RAG and evals — because in 2026 this is core engineering, not a side quest.
Practised, not just read
Every concept is exercised: the Prompt Lab runs and scores your prompts, and you build and operate your own agent (Mysty) on your free cloud.
Grounded in what breaks
Hallucination, prompt injection, data leakage, runaway cost — you learn the failure modes and the guardrails, tied to real incidents.
Six modules, from tokens to production
A build-first path from how models work to shipping governed AI tools your company can rely on.
LLM foundations
How large language models actually work: tokens, context windows, temperature, cost and latency trade-offs, and where models fail — so you reason about them, not cargo-cult them.
Prompt engineering
COSTAR and canonical technique — role framing, few-shot, structured output, delimiters, chain-of-thought and anti-hallucination — practised live in the Prompt Lab with scored rewrites.
RAG done right
Retrieval-augmented generation end to end: chunking, embeddings, vector stores, hybrid search, re-ranking, grounding and citations — and how to stop it hallucinating.
Agents & tool use
Planning, tool/function calling, memory, multi-step loops and guardrails. Build agents that do real work safely — and know when an agent is the wrong tool.
Evals & observability
You can't ship what you can't measure: build eval sets, judge outputs, catch regressions, trace tokens and cost, and keep quality from drifting in production.
Company AI tooling
Build the internal tools companies run: retrieval over private docs, gateways and key management, MCP tool servers, and putting a human safely in the loop.
Four things you'll actually build
Every module ends in something that runs — not a quiz.
Score a prompt
Write a prompt, watch it run against a real model, and get scored on COSTAR with an optimized rewrite and rationale.
Ship a RAG answer
Chunk and embed a corpus, retrieve, ground the answer with citations, then evaluate whether it actually improved.
Run an agent
Give an agent tools and a goal, add guardrails and memory, and watch it plan and execute — safely and observably.
Prove it with evals
Build an eval set, judge outputs at scale, and catch the regression before your users do.