Accuracy of the information retrieval pipeline plays a key role in the quality of Retrieval Augmented Generation (RAG) applications. GraphRAG is one of the advanced RAG techniques that improves accuracy of the information retrieval using graph-like relationships in your data. This solution accelerator shows you how to build a legal research copilot using this advanced technique on top of Azure Database for PostgreSQL.
This solution accelerator integrates GraphRAG with PostgreSQL graph query capabilities, providing a complete GraphRAG platform with a minimal number of platform components. The solution accelerator uses a semantic ranker for relevance boosting, the Apache AGE extension for graph queries, and the azure_ai extension for integration with Azure AI Services, including Azure OpenAI. Learn more.