In this Community Call, we are joined by Graph.Build. They'll take the 2023 Amazon Reviews dataset - 571M+ reviews with second-level timestamps, rich item metadata (price, images, attributes) and “bought-together” link graphs - and show the steps of turning it into a governed property knowledge graph.

With a GraphRAG layer on top, you can ask questions like: “Why did ratings drop for Model X in Sept 2023?” or “Which accessories are most often co-purchased by similar cohorts, and what explanations do power users give?”.

What we’ll cover / what you’ll learn:

Model & Build: Designing an enterprise-ready graph model (users ↔ reviews ↔ items ↔ ‘bought-together’ links), mapping semi-structured and structured sources in Graph.Build

Schema: Generating Memgraph-ready schema and ingesting into database

GraphRAG in Practice: How to set up a GraphRAG pipeline on Memgraph on top of the Amazon Reviews dataset

How to Build an Amazon-Scale Knowledge Graph for GraphRAG