Founding batch: waitlist open
Production AI Engineering
Go from calling an LLM API to designing, evaluating and deploying agentic systems that hold up in production. Built for developers who already ship code.
- Who it's for
- Working developers with 2–8 years of experience
- Format
- Recorded lessons released weekly, plus a live Q&A every week
- You finish with
- A production-grade AI system in your portfolio
What you'll learn
Module 01
Foundations that matter in production
Model selection, structured outputs, tool calling, context windows and why prompts are code.
Module 02
Harness engineering
The scaffolding around the model: loops, state, retries, tool routing and memory. Where most of the real engineering lives.
Module 03
Agents and multi-step workflows
When to use an agent and when a pipeline is better. Planning, delegation and keeping agents on track.
Module 04
Retrieval that holds up
Chunking, hybrid search, reranking and grounding answers in real company knowledge.
Module 05
Evals
Building test sets, LLM-as-judge done right, regression checks in CI and measuring what users actually care about.
Module 06
Guardrails and security
Prompt injection, permissions, PII, human-in-the-loop and audit trails.
Module 07
Cost, latency and scale
Caching, model routing, streaming, batching and keeping the bill predictable.
Module 08
Shipping it on AWS
Serverless deployment, observability, secrets and the boring parts that make it production.
This is for you if…
- ✓You have 2–8 years of experience and ship code for a living.
- ✓You've built a chatbot or a RAG demo and want to go much further.
- ✓You want to move into an AI engineer role, or lead AI work in your current one.
This is not for you if…
- ✕You're looking for a beginner programming course.
- ✕You want to train or fine-tune models. This is about building systems around them.
- ✕You want theory without code.
Get founding-batch pricing
The first batch is small and priced lower in exchange for feedback. Join the waitlist and you'll hear about dates and pricing before anyone else.