Inspiration What it does How we built it Challenges we ran into Accomplishments that we're proud of What we learned What's next for Graph_RAG_supplychain This project is an AI-powered GraphRAG-based supply chain intelligence system designed to analyze and predict semiconductor and GPU supply chain risks in real time. By integrating Graph-based Retrieval-Augmented Generation (GraphRAG) with GPU-accelerated graph analytics, it allows businesses to make data-driven decisions using natural language queries.
The system stores, processes, and visualizes complex trade relationships using ArangoDB, NetworkX, and NVIDIA cuGraph while leveraging LLMs from Hugging Face and OpenAI to generate AI-driven insights. Users can query supply chain dependencies, identify risks, and explore alternative supplier networks, all through an interactive AI-powered assistant.
This solution is designed to help companies like NVIDIA navigate geopolitical risks, trade restrictions, and material shortages, ensuring resilient and optimized supply chain operations.