Kùzu - A Fast, Scalable Graph Database for Analytical Workloads

Learn about a highly scalable, fast, and user-friendly open-source embedded graph database designed for analytical query workloads in this comprehensive conference talk. Explore state-of-the-art methods from graph database research while discovering how to transform relational datasets into knowledge graphs. Follow along with practical demonstrations that showcase running Cypher queries, analyzing datasets using graph algorithms, and training graph neural networks with PyTorch Geometric to compute node embeddings. Master techniques for storing embeddings in the graph database and understand how these methods can enhance RAG systems integrated with LLMs. Gain insights into core innovations and distinctive features that set this database apart from other graph database solutions through hands-on examples and detailed technical explanations.