I have spent the last few months building AI agents with LangChain and LangGraph, and one thing kept bothering me: most tutorials stop at a demo that breaks the moment you touch it. So I decided to build the course I could not find.
AI Agent Engineering opens on LogicStack on 1 November 2026. It is practical and project-first:
- Module 0: Orientation and setup, your first LLM call
- Module 1: LangChain core (runnables, parsers, memory)
- Module 2: Tools and agents (ReAct, error handling, retries)
- Module 3: RAG that answers from your documents, with citations
- Module 4: LangGraph foundations (state, routing, interrupts)
- Module 5: Advanced LangGraph (checkpointing, human-in-the-loop, multi-agent patterns)
- Module 6: Production (evaluation, guardrails, FastAPI and Docker deployment, EU data rules)
- Capstone and final exam
7 modules, 39 lessons, a working project at the end of every module.
One deliberate choice: the certificate is not issued for finishing the videos. You get it only after you pass the final exam. A certificate should mean somebody checked the work.
If you are learning agents right now, I would genuinely like to know: what is the one topic you wish courses covered properly? I am still finalising the last lessons, and I will read every answer.
Course page: https://logicstack.org/courses/ai-agent-engineering
