Just Add an LLM: Creating a New Food Recommendation Engine From an Old Cypher-Query Based One

Session Track: Data Science

Session Time: 10:30 pm - 11:00 pm November 7

Session description

This session revisits a food recommendation tool that helps people make healthier food choices. In the session, you will learn how to upgrade a system that generates personalized recommendations based on Cypher queries to an AI-supported recommendation engine that incorporates a Large Language Model (LLM). The session focuses on comparing two approaches to building AI-supported recommendation engines. The baseline approach uses hard-coded Cypher queries. The additional approaches include using LLM-generated Cypher queries based on a prompt, and an LLM using Retrieval-Augmented Generation (RAG). Both approaches rely on the same eating events graph to serve the recommendations.