AI GitHub
Microsoft GraphRAG for synthesizing a corpus
Build entity relationships and summaries that support questions across many documents.
Key Takeaways
- The indexing pipeline extracts entities and relationships, detects communities and generates summaries.
- Set up the model connections and plan an indexing budget before importing a full document library.
- Start with a small corpus about several connected events.
- Indexing calls and generated intermediate artifacts can dominate cost.
How it works
The indexing pipeline extracts entities and relationships, detects communities and generates summaries. Query modes can use local evidence or higher-level community reports.
What you need
Set up the model connections and plan an indexing budget before importing a full document library. Keep generated graph and summary artifacts available for inspection.
What to try first
Start with a small corpus about several connected events. Compare a broad thematic question with one that needs an exact source sentence.
Ask a question that combines entities from separate communities. Check which source documents still support the synthesized answer after intermediate summaries have compressed them.
Where to look for failures
Indexing calls and generated intermediate artifacts can dominate cost. Refresh and audit those artifacts when the underlying documents change.
Related Franklin pages
Use a question that needs evidence from several sources and another that needs an exact quotation. This helps show when building intermediate relationships or summaries is worth the effort.
- Turn inputs into connected knowledge and retrieve that context for AI applications. Cognee for building memory from documents
- Study a research pipeline that generates questions, gathers sources and drafts a structured article. STORM for researching and drafting cited articles
For the full path from source documents to generated answers, start with Franklin’s glossary. Retrieval-augmented generation
Project documentation
Try it for your workflow
Explore the project’s documentation for setup and usage, or save this page to a collection for later.
Read the project documentation ↗