FileFlux
A .NET library that turns PDF, DOCX, HWP and other documents into optimized chunks for RAG systems, through a five-stage pipeline built on high-performance Rust FFI readers.
FileFlux is a .NET document-processing library that transforms PDF, DOCX, XLSX, PPTX, HWP, HWPX, Markdown, HTML, TXT, JSON and CSV documents into chunks optimized for retrieval-augmented generation. A five-stage stateful pipeline — extract, rule-refine, LLM-refine, chunk, enrich — runs through Rust FFI-based native readers that the project reports as two to five times faster than the alternatives, with pinned NuGet binaries for reproducibility and an opt-in runtime self-update.
Chunking strategies include automatic, smart, intelligent, semantic, paragraph, fixed-size, hierarchical and page-level modes, and chunks carry structural metadata such as heading path, page numbers and context-dependency scores, plus a document graph tracking sequential, hierarchical and semantic relationships between chunks. AI service interfaces are defined separately from their implementation, so a preferred provider can be plugged in, and a smaller FileFlux.Core package offers extraction only, without chunking, for minimal-dependency use.
It installs via NuGet and is MIT licensed.