Stop chunking tables: How we built an agentic GraphRAG for financial disclosures with Docling

This blog post discusses an innovative approach to improving retrieval-augmented generation (RAG) pipelines for handling SEC filings, specifically focusing on the limitations of typical text chunking methods. The author outlines a new open-source stack called 'agentic-graphrag-finance' that leverages Docling to parse financial documents into structured knowledge graphs, enhancing accuracy in financial Q&A systems by maintaining relationships between data elements and their context.