AI engineer with TypeScript
The same AI engineering, in the language of the web.
At the end: Build LLM features, RAG and agents in TypeScript and ship them inside a real web backend.
Stage 1: Foundations
The language and the model.
- JavaScript, inside out: How JavaScript really runs: scope, closures, the event loop.
- TypeScript for real code: Types that catch the bug before it ships.
- Building with LLMs (TypeScript): Messages, tokens, streaming, tools and zod output, on the raw API.
Stage 2: Core AI engineering
The things every AI product is made of.
- RAG: search that answers: Search that answers, with Postgres and pgvector.
- Agents from scratch (TypeScript): The agent loop from scratch, with zod tools.
- MCP servers and clients in TypeScript: Give any assistant your tools through the Model Context Protocol.
- Evals that catch regressions in TypeScript: Catch a regression before your users do.
- Memory and multi-agent systems in TypeScript: Remember across sessions and split work between agents.
- Harness engineering (TypeScript): Everything around the model that makes an agent reliable.
- AI security: attack and defend (TypeScript): Break your own agent, then defend it, before an attacker does.
- Build your own coding agent (TypeScript): Build the coding agent itself: tools, safe edits, a sandbox and a loop that fixes.
- The Vercel AI SDK: Ship a streaming AI chat in React with the Vercel AI SDK.
- Claude and the Anthropic SDK (TypeScript): Claude the right way: tools, caching, thinking and batches.
- Build an AI gateway (TypeScript): Run AI in production: one gateway over many providers, within budget and SLO.
- Fine-tuning basics (TypeScript): When to fine-tune, and how: LoRA on a model you can see inside.
- Build a voice agent (TypeScript): A phone agent that listens, answers in time and knows when to stop talking.
- AI engineering traps (TypeScript): The AI bugs interviews ask about, found and fixed.
Stage 3: Build it for real
Whole AI apps for a company, in TypeScript.
- AI support copilot project (TypeScript): A support copilot that answers from policy.
- AI booking agent project (TypeScript): A booking agent that acts, safely.
- AI invoice reader project (TypeScript): Read invoices into checked, structured data.
- MCP server project (TypeScript): A production MCP server.
- AI answer grader project (TypeScript): An AI grader you can trust.
- Build your own GPT project (TypeScript): Build a small GPT and see how the model works inside.
Stage 4: Interview ready
The bugs interviewers ask about.
- TypeScript traps: TypeScript bugs the compiler lets through.
- JavaScript traps: JavaScript bugs from production.