Live sessions on agents, retrieval, benchmarks, and building in public. History class meets lab notes.
Sparse retrieval architecture for production hybrid search. How MiniCoil closes the gap between dense and sparse without the compute overhead.
How retrieval becomes the memory layer for agents: grounding, provenance, and query-time tradeoffs.
Live demo: building search-enabled agent systems from scratch.
Production reliability, fault tolerance, and what it actually takes to run vector search at scale.
How graph structure and vector retrieval combine for knowledge-heavy applications.
Interview sur l'IA, la recherche vectorielle, et le futur du travail avec les agents.
Persistent storage, caching strategies, checkpointing. Led to GKE Data Cache feature.
199 attendees. Evaluation methodology for embedding model selection in production.
Hands-on training session for production retrieval systems.
RAG-enabled agents for complex queries across multiple data modalities.
Platform powering 2M+ conversations across three countries.
AI security, vector databases, and agent reliability for production systems.