A common UX challenge with RAG applications is that they often leave users staring at a blank page, unsure of what queries to run to explore a dataset. This can be overwhelming and inefficient.

GraphRAG transforms this experience with its entity-based query generation feature. By combining structured data (entities and relationships) with unstructured data (community reports and covariates), it crafts insightful follow-up queries tailored to a session's query history, highlighting critical themes and information.

With GraphRAG, users can be guided to uncover and explore insights from datasets.

Try out the live demo showcasing GraphRAG's potential in query generation: https://graphrag-demo.deepset.ai/query-gen