Zach Blumenthal

Zach Blumenthal is an AI Engineer community speaker whose work centers on GraphRAG: combining knowledge graphs, semantic search, and graph data science to make generative AI applications more contextually useful.

At AI Engineer World’s Fair 2024, he helped lead a hands-on GraphRAG workshop that built a retail recommendation system using product-description embeddings, Cypher queries, collaborative filtering, and customer purchasing histories. The application personalized identical searches for different customers by combining semantic relevance with signals from shoppers who bought similar products.

The workshop also used node embeddings and nearest-neighbor analysis to identify likely co-purchases, then incorporated those recommendations into customer-specific marketing emails. Blumenthal addressed practical implementation tradeoffs, including sparse purchase histories, graph-first versus vector-first retrieval, recommendation scoring, and recalculating graph embeddings as underlying data changes.