plugin-docpixie
v1.1.1
Published
Adaptive RAG agent for document analysis — OCR + LLM Vision + structured extraction.
Maintainers
Readme
plugin-docpixie
Overview
Adaptive RAG agent for document analysis - OCR + LLM Vision + structured extraction.
Features
- Advanced Document Processing: Uses visual LLMs to understand complex document layouts (tables, images, charts).
- Adaptive Chunking: Intelligently chunks documents based on semantic structure rather than just token count.
- Structured Extraction: Can pull specific JSON schemas directly from scanned PDFs or images.
Usage
- Enable the plugin.
- When creating an AI Employee or Knowledge Base ingestion pipeline, select DocPixie as the extraction engine.
- Upload complex documents (PDFs, images).
- The agent will process them, maintaining spatial and structural context for better RAG answers.
