Parsing documents for air-gapped RAG: no cloud, no JVM, no Python

Short summary

DeepDoc is a single static Rust binary that parses 13 born-digital formats (docx, pptx, xlsx, pdf, etc.) into clean Markdown, JSON, or plain text with zero runtime dependencies and no network access. It targets air-gapped RAG pipelines where cloud parsers like LlamaParse are disqualified and JVM or Python stacks are overkill. The tool preserves document structure for smarter chunking and produces deterministic output for reproducible indexes.

• •Single Rust binary parses 13 formats to Markdown with no JVM, Python, or network
• •Designed for air-gapped environments where documents cannot leave the machine
• •Preserves structural metadata for structure-aware chunking in RAG pipelines

Generated with AI, which can make mistakes.