From embedding to production. Design, orchestrate and deploy robust RAG systems: hybrid retrieval, LangChain / LlamaIndex chains, evaluation and monitoring in real conditions.
Ready to industrialise your RAG systems?
RAG (Retrieval-Augmented Generation) lets a language model answer from your own documents and databases: the system first retrieves the relevant passages, then the LLM generates its answer based on them.
Yes, 100% online: 5 days (35 hours) of live sessions, hands-on labs and replays. You leave with the notebooks, the code repositories of your final project and a certificate.
Intermediate to advanced. The course is taught in Python, with LangChain, LlamaIndex and FastAPI.
LangChain, LlamaIndex, the OpenAI and Anthropic APIs, the Pinecone, Qdrant and pgvector vector databases, FastAPI, Prefect for orchestration, Ragas and LangSmith for evaluation.
RAG gives an LLM access to your knowledge; an AI agent plans and carries out actions in your tools. The two are often combined: see the AI Agents & MCP course.
Jobseekers registered in France can apply for AIF funding from France Travail. Companies can fund it through their skills development plan.