LLM Zoomcamp is a free online course that teaches how to build production-ready LLM applications using retrieval-augmented generation, vector search, embeddings, and related techniques.
The course addresses the challenge of moving from LLM fundamentals to deployed systems that can answer questions over custom knowledge bases. It takes a hands-on approach, structuring learning around practical implementation over ten weeks. Participants build working codebases while learning retrieval-augmented generation, vector search, embeddings, AI agents, function calling, evaluation, monitoring, hybrid search, and reranking.
The course suits software engineers adding LLM capabilities to products, data engineers working with vector search and retrieval pipelines in production systems, and ML practitioners needing structured approaches to evaluate and monitor LLM applications. It requires only Python proficiency, basic command-line comfort, and Docker familiarity; no prior ML or LLM experience is necessary, and no GPU is required. Participants can engage either through live cohorts with graded homework, peer review, and certificates, or self-paced learning with access to all materials without scoring or certification.
The project maintains active development with regular course cohorts, documented logistics and curriculum materials, video lectures organized in playlists, and a community Slack channel for participant discussion. Course projects from participants are publicly showcased, and the repository receives updates to reflect new modules and cohort schedules.
