NODES AI 2026 - The Searchable Grid: Building a GraphRAG-Powered Digital Twin for Real-Time Insights

This session features Lijimol George, Boobalan Sivaprakasam, and Mathumitha G S discussing "The Searchable Grid: Building a GraphRAG-Powered Digital Twin for Real-Time Insights". The talk addresses the challenge utilities face in managing the vast data produced by electrical networks, such as GIS, AMI telemetry, and transformer health, which are often trapped in static diagrams and siloed systems. The speaker presents a solution utilizing Neo4j Knowledge Graphs alongside GraphRAG to construct a dynamic Digital Twin of the distribution grid, enabling engineers to perform complex queries in natural language and receive contextual, relevant responses. Attendees will gain insights into fault tracing, outage prediction, and grid optimization through an integrated approach that combines graph technology with AI capabilities.