AI with Angular
The free, written version of the workshop. Two tracks, because "AI with Angular" is two problems: coding agents working inside your codebase, and an LLM working inside your product. Each track has its own opener and its own short parts. One closing part serves both.
"AI with " hides two separate problems, and the first mistake most teams make is treating them as one.
In the first, the AI builds your Angular: a in the repository, and the danger is to the codebase. In the second, the AI runs inside your Angular: a model behind a feature, and the danger is to the product. Different risks, different guardrails, different people in the room. So the series is two tracks. Read the one you need. Each part is short, has one lesson, one prompt you can paste into your own agent, and a link to the exercise in the demo repo. Everything runs on Atlas Bank, an imaginary Moroccan bank, never client code.
Agents in your codebase
Copilot, ChatGPT and Codex, Claude Code or Cursor writing and refactoring code in your repository. The risk is to the codebase: eroded architecture, crossed boundaries, tests rewritten to pass. The fix is a small set of files the agent must read, pass and stay inside, and a migration discipline that survives production.
Coming soon
- Same AI, two outcomes: a faster team or a dead project
- The three files an Angular monorepo needs before an agent writes a line
- From ticket to task file: the agent drafts, a person decides
- Pin the behaviour before you touch it
- Demote, don't delete: the migration plan that survives production
- What could I delete and stay green?
AI inside your product
An LLM behind a feature: an assistant, a summariser, a form that fills itself, a screen the model can drive. The risk is to the product: latency nobody planned, cost that grows with users, data the model should never have seen, a demo that never becomes a feature. The fix is the plumbing the demo skipped.
Coming soon
- The demo takes twenty minutes. The feature takes two months.
- Streaming into signals, with a Stop button that stops the bill
- Tool calling, and the cost nobody demos
- The data the model may see
- Agentic UI: real or demo?
The closing part
Coming soon
- What do we forbid? The team policy, and the numbers that prove it worked
Fast and safe: AI in your Angular codebase
This series is the written version of the workshop. In the room you do the same thing on your own codebase, with your team, in 3 days (or 6 half-days remote).
See the agenda