I build retrieval and memory systems for agents: hybrid search, eval harnesses, and production demos.
Chrome DevTools for AI agents: trace viewer, attribution overlay, diff mode, critical path analysis, portable repro bundles.
Retrieval evaluation + tuning workflow: ingest, index variants, eval suite, recommendations. Built for repeatability.
NBA stats, Balatro, CAMB, and the cost of curiosity.
From Q&A chatbots to systems that plan, retrieve, act, and verify.
How native MMR breaks the echo chamber of similar results.
Persistent storage, caching strategies, checkpointing. Led to GKE Data Cache feature.
Sparse retrieval architecture for production hybrid search.
199 attendees. Evaluation methodology for embedding model selection.
Hands-on training session.
RAG-enabled agents for complex queries across multiple data modalities.
Platform powering 2M+ conversations across three countries.
Practical strategies for prototype to production.
AI security, vector databases, and agent reliability.
I work on retrieval, evaluation, and agent reliability at Qdrant. I help teams ship production-grade semantic search, hybrid retrieval (dense + sparse), and agentic RAG pipelines with clear latency, cost, and quality tradeoffs.
Previously: ML systems for the Department of Homeland Security at SAIC (classification + forecasting). English and French. University of Maryland, College Park.
Fastest on X. Open to talks, consulting, and collaborations. San Francisco, PT.