ocr-vision

An elegant, production-ready AI document assistant that uses vision-based OCR to transform PDFs and images into structured markdown notes, summaries, flashcards, and quizzes. Features a rich block editor, document chat, and automatic Mermaid diagram generation.

Transform any document, slide, or handwritten note into structured study guides, podcasts, and Anki decks instantly. Powered by Gemini, Groq, and ChromaDB.

ML/AI workers for Innova EdTech. Python 3.11 + Anthropic Claude + Gemini Vision. Implements BKT calibration, IRT fitting, LLM error classification, and OCR vision pipeline. AWS Lambda container images.