// learn · backend
Backend engineer to AI engineer
You already ship APIs, services, and databases. AI engineering is the same craft with one new kind of component — a model — and a few patterns around it. Here's how what you know maps onto what you'll build.
Start with the guide this path is built around: Backend engineer to AI engineer.
Nothing here is a reset. Each thing you already do has a direct AI analog.
What you know
REST / gRPC APIs and services
What you'll build with it
Model calls treated as just another dependency — with timeouts, retries, and fallbacks around a non-deterministic endpoint.
What you know
Databases and query layers
What you'll build with it
Vector search and retrieval — pgvector alongside the Postgres you already run, powering a RAG service.
What you know
Background jobs and queues
What you'll build with it
Tool-using agents that call your functions, with the same idempotency and error handling you already enforce.
What you know
Auth, rate limiting, and caching
What you'll build with it
Guardrails, token budgets, and response caching that keep an LLM feature affordable and safe under real traffic.
Work these in order. Every link is free to read.
- 01The AI Engineer Roadmap
The six-stage path from concept to offer — read it first as your map.
- 02RAG in production
Retrieval-augmented generation: grounding a model in your data. The pattern behind most useful features.
- 03Agentic AI
Tools, memory, and guardrails — models that call your functions, with the reliability you already enforce.
- 04Build a production RAG app
A full end-to-end build where your API, database, and caching skills transfer directly.
- 05Interview prep
Rehearse the RAG and system-design questions AI teams actually ask.
You have the engineering foundation. Point it at models.
Production AI Notes
One practical AI engineering email each week
One concept, one architecture, one project idea, and one interview question — written for developers who want to build and ship real AI systems.