Retrieval Layer in Context Engineering: From Intuition to Production

Context engineering is the art of orchestrating the information your AI systems have at their disposal. AI agents' knowledge is only as good as the knowledge they can access, and how you build that information retrieval layer determines everything downstream. This workshop approaches context engineering from a vector search standpoint, covering the full range of tooling it enables (recommendations, hybrid search): where it genuinely fits and where it doesn't, and then what & how to use instead: knowledge graphs in GraphRAG. We'll go from the intuition behind context engineering to context engineering in practice, through a real application: a medical AI research copilot. You'll learn the full picture: how to increase search relevance and reduce retrieval latency, best practices for prompt engineering for context retrieval, and how to orchestrate vector-search-backed context-engineering tools. By the end of the workshop, you'll be able to build a context-engineering layer for a simple medical AI copilot.