GraphRAG Inference Hackathon by TigerGraph

Welcome to the GraphRAG Inference Hackathon by TigerGraph! A beginner-friendly online coding challenge where you'll prove that graphs make LLM inference faster, cheaper, and smarter. About the Opportunity: LLMs are powerful, but they're also expensive and slow. Every time you ask them something complex, they burn through thousands of tokens trying to reason their way to an answer - and at scale, that cost adds up fast. Graphs offer a smarter path. By organizing information into relationships the model can actually follow, graphs help LLMs focus on what matters - cutting tokens, speeding up responses, and saving cost, all without losing accuracy. The GraphRAG Inference Hackathon by TigerGraph is your chance to prove this with real numbers. You will design, build, and benchmark a working product that shows exactly how much better inference gets when graphs enter the picture. New to GraphRAG? No problem. This hackathon is designed to be beginner-friendly - you don't need prior experience with TigerGraph or graph databases to participate. If you know Python and have played around with LLMs or APIs, you're ready to build. We'll provide documentation, starter resources, and mentoring sessions for the Top 10 teams so you have everything you need to ship a great project. Whether you're a student exploring GenAI for the first time, a developer curious about graphs, or an AI engineer building production RAG systems, this hackathon is your playground to learn, build, and compete. What You'll Build: Your mission is to build a working product with two parallel pipelines answering the same question, plus a dashboard that compares them head-to-head. Pipeline 1: The Baseline (Just LLM): You send a prompt, the LLM answers. Simple, direct, but expensive and often slower than it needs to be. This is your control group. Pipeline 2: GraphRAG (Graph + LLM): Your prompt first runs through TigerGraph, which pulls the most relevant context and maps out the connections between entities using multi-hop reasoning. The LLM then generates its final answer using this filtered, structured context - not the entire document dump. The Comparison Dashboard: A side-by-side scoreboard that automatically tracks tokens used, response time, cost per query, and answer accuracy - proving where GraphRAG wins. How to Structure Your Solution: Follow the AI Factory model - a clean, production-ready architecture with four separate layers: Graph Layer - TigerGraph handles entity extraction, relationships, and graph queries. Inference Orchestration Layer - decides when to use graph, when to call the LLM, and how they work together. LLM Layer - generates the final answer using the filtered context. Evaluation Layer - runs the benchmarks and populates your comparison dashboard. Keeping these layers separate makes your solution scalable, reusable, and ready for real-world production - exactly what the judges are looking for. Eligibility and Registration: Open to students and working professionals globally - Beginners are welcome! Team size: 1 to 5 members (solo participation allowed). Each participant can be part of only one team. Cross-institutional teams are welcome. Participation is completely free. Rewards and Prizes: Winner: 1st Place: $250 cash prize, certificate, and recognition for building the best GraphRAG inference system. 1st Runner-Up: 2nd Place: $150 cash prize and certificate for outstanding GraphRAG implementation. 2nd Runner-Up: 3rd Place: $100 cash prize and certificate for strong graph utilization and efficient inference design. TigerGraph Community Leads' Exclusive Winner: $100 cash prize and certificate — a special award for creativity, innovation, and community spirit. All participants who submit valid solutions will receive a digital certificate of participation. Join the Hackathon WhatsApp Group: Connect with fellow participants, find teammates, ask questions, and stay updated on all hackathon announcements by joining our official WhatsApp group: https://chat.whatsapp.com/Iwdyhie2gSoIR0k2teMtKb This is where all important updates, resources, and mentor interactions will happen - make sure to join right after registering!