Building Real-Time Recommendation Systems with Neo4j

Session Track: Data Intelligence

Session Time: 12:00 am - 12:30 am November 7

Session description

Building recommendation engines is hard. You need to consider how to generate and rank recommendations, meet technical constraints like low latency, and plan a strategy to ensure a smooth deployment. In this session, Milan will show you how to build a recommendation engine that generates real-time recommendations for an event-driven email sending system. Milan will go over how to upload user and business data to Neo4j, leverage custom vector indexes for vector search, create a cypher query to generate and rank recommendations based on similar users, and lastly deploy to a lambda function on AWS. With a practical, enterprise use case, this session aims to show how Neo4j is able to generate high-quality recommendations even in low-latency conditions.