Learn with Neo4j's New "Get to Know Graph & GenAI" Webinar Series >>
Session Track: Knowledge Graphs
Session Time:
Session description
Behind every hiring pipeline lurks a maze of SQL tables that still power clumsy keyword search or GPU-hungry, fine-tuned models. This session shows how shifting that same data into Neo4j—and pairing Cypher reasoning with its native vector index—unlocks a graph-native RAG job matcher that surfaces the best candidates in milliseconds, explains each recommendation in plain language, and costs pennies per thousand queries. Attendees will see how the graph captures nuanced relationships like “must-have” versus “nice-to-have” skills, how hybrid retrieval (graph paths + embeddings) boosts Hit@5 against a fine-tuned BERT baseline, and how the approach scales to hundreds of thousands of profiles without ever touching model weights. Walk away with a clear blueprint for turning any set of relational HR tables into an explainable, low-cost talent-matching service—no expensive GPUs or no black-box magic, just Neo4j and a dash of LangChain.
ML Solutions Architect, X-Team
Otávio Calaça Xavier has 20 years of experience in web applications development and 13 years as a professor in Computer Science undergraduate courses. He has participated as a speaker in more than 100 events around the country (Brazil). Currently, he is an ML Engineer at X-Team, a professor at UFG (one of the top 40 LATAM universities) and a Ph.D. student in Computer Science, with graph neural networks and RAG being his main areas of study.