Built a tool solve the nightmare of chunking tables in PDF vs. Markdown
Stop using static chunk sizes. A lightweight, production-ready RAG ingestion toolkit. Uses Docling for layout-aware parsing and applies smart heuristics for optimal chunking (PDF vs Code vs MD). Extracted from a production RAG platform. After years of frustration with how LLMs handle complex documents, especially PDFs with tables, I decided to build a solution myself.
Open source
69 stars · GitHubexternal siteArchived
Reactions (1)
Similar products
by description similarity
RAG-chunk
github.comA CLI to test RAG chunking strategies
Chonkie
github.comA Fast, Lightweight Text Chunking Library for RAG
AutoRAG
github.comAutoML tool for RAG. Find an optimal RAG pipeline
RAGBuilder
github.comHyperparameter tuning on various RAG parameters
FastGraphRAG
github.comBetter RAG using good old PageRank