Become an AI Engineer
You know how to use AI to build a product. Now learn how to build AI into the product.
Five intensive sessions with one fully functional AI-enabled product built that can:
- Use an LLM behind the scenes
- augment it with specific domain knowledge
- and custom MCPs, tools and prompting
- within a managed context
- that can scale with the task
This cohort is closed now
Please register interest so that we can let you know when there is a future cohort available.
Who is this for?
Anyone who has built a product, by coding it or with AI tools like Claude Code or Codex, and now wants to build AI into one.
Engineers go deep on evals, retrieval, tools, context and security. Production practice, not prompt tips.
Vibe-coders, PMs, students, graduates or other professionals lean on AI-generated code, get help every session, and plan extra homework time.
What you'll need
- A laptop with Python and your own AI coding tool
- Time for 5–8h of homework a week
Not for: complete beginners, or anyone after ML theory or model training.
Format
- 10:00–11:00Review homework and Q&A
- 11:00–12:00Introduce the session's topics, with coding examples
- 12:00–12:30Lunch break, food provided
- 12:30–16:00Work on your project
- 5 Sunday 10:00–16:00, in person (1, 8, 15, 22 and 29 November 2026)
- Finish coding the feature as homework
- To make things faster and focus on what's important, lean on AI tools for coding
- Missing one session has an impact, but there are ways to deal with it; missing more than one is not recommended, as it may significantly impact the outcome
This cohort is closed now
Please register interest so that we can let you know when there is a future cohort available.
Learn from someone who's made the move.
A software developer with 25 years of experience who has already transitioned into AI engineering.
Foundations
Project
- Defining the project we'll all work on
- Setting expectations and defining what is essential
- Setting up the tools that we'll need throughout
- Protecting against untrusted user input
LLM API
- Understanding the LLM APIs that are available
- Making calls and stacking up messages from a conversation
Testing
- The eval suite as the feature's regression tests
- Prompt optimisation using automated tests
Retrieval (RAG)
Core
- Adding a document with domain knowledge using embeddings
- Using a text embeddings model to create the embeddings
- Storing and querying embeddings in PostgreSQL pgvector
- Prompting the LLM to give references for its answers
Testing
- Retrieval regression tests (query → expected chunk)
- E2E golden set
Security
- Indirect injection via retrieved docs
- PII in corpus
Extras
- Chunking strategies, hybrid search, reranking, failure modes
- When not to RAG
Deliberately light: it gives Iterators with a difficult start in week 1 time to catch up.
MCPs & Tools
Core
- Understand tool calling
- ReAct loops with tool calling
- Create a custom MCP for your tools
Testing
- Tool-call assertions with mocked tools
- Loop-termination tests
Security
- Least-privilege tools
- Argument validation
- Confirmations for dangerous actions
- MCP trust boundaries
Extras
- Planning patterns
- Guardrails
- When agents are overkill
Context management
Core — compaction
- Context as a capped budget, triggers, keep-zones
- Summarization vs. pruning/masking
- Context drift
- Dropped-constraint failure modes
Testing
- Constraint-preservation tests through compaction
- Full regression suite
Security
- Compaction must not drop safety constraints
Extras
- Rewind/undo: checkpointing, conversation rollback, regeneration, agent-state recovery
- Long-term memory: user-scoped cross-session memory; minimal in-house build, Mem0 mentioned as an off-the-shelf shortcut
- Memory testing and security: recall tests, PII in memory, cross-user scoping, memory poisoning
Advanced topics + demo day
10:00–12:00 · Presentation
- Cost management: token accounting, caching economics, model routing, cost caps
- Distributed/specialized agents: agents with their own context and specialty, context isolation, when to split
- Hardening checklist: rate limits, abuse, PII in logs/traces, key review
12:00–16:00 · Demo day
- Brownbag demos
- Completion feedback
Principles
- Learn by doing, not from a presentation
- Everyone moves at their own pace
- Students can choose what they want to focus on
This cohort is closed now
Please register interest so that we can let you know when there is a future cohort available.