# RAG · Tech watch

[Canonical source](<https://sebastien.rigaux.be/en/tech-watch/tags/rag>)

Augmenting model responses by retrieving relevant documents and context.

## [RAG-Anything: handling charts in multimodal documents](<https://sebastien.rigaux.be/en/tech-watch/links/rag-anything-handling-charts-in-multimodal-documents>)

RAG-Anything is a framework built on LightRAG for indexing and querying documents that mix text, images, tables, equations, and charts. It coordinates parsing and content analysis in a single multimodal pipeline.

The project combines vector similarity with a knowledge graph, provides processors for different content types, and can pass retrieved images to a vision model during queries. A recent change also routes rendered charts through the image pipeline while preserving captions and extracted text. To update a previously indexed document, multimodal reprocessing must be explicitly enabled with `force_multimodal_reprocess=True`.

2026-10-29 · [Markdown](<https://sebastien.rigaux.be/en/tech-watch/links/rag-anything-handling-charts-in-multimodal-documents.md>) · [Original resource](<https://github.com/HKUDS/RAG-Anything>)

## [SurfSense: work on documents with a local agent](<https://sebastien.rigaux.be/en/tech-watch/links/surfsense-work-on-documents-with-a-local-agent>)

SurfSense provides a document application whose index stays on the machine. Choosing a local or remote model lets users adapt processing to document privacy requirements.

2025-05-01 · [Markdown](<https://sebastien.rigaux.be/en/tech-watch/links/surfsense-work-on-documents-with-a-local-agent.md>) · [Original resource](<https://github.com/MODSetter/SurfSense>)