How aryanahamed/handwritten-to-digital-notes works
Project type and purpose
• This is a single‐service, full‐stack web application built on Streamlit (a Python framework) that ingests handwritten notes, transcribes and “cleans them up,” enriches them via a large language model, and delivers a downloadable PDF.
• The README calls out two main external services: the Groq API (for vector/OCR processing) and the Llama 3.2 LLM (for rewriting, explaining, and formatting).
File structure
• Top level:
• main.py – the entire Streamlit application, handling UI, input, business logic, and output.
• requirements.txt – lists Python dependencies (streamlit, groq SDK, llama3.2 client, PDF library, etc.).
• README.md, LICENSE, .gitignore, .gitattributes.
• There are no separate frontend/backend folders. The Streamlit script combines both layers.
• No database or persistent storage directories – this is an in‑memory, stateless service.
README for additional insights
• Key features:
• Upload handwritten notes → extract topics + text
• Support for LaTeX, code blocks, tables
• PDF download with enriched content
• External dependencies: Groq API & Llama 3.2 LLM.
• Hosted as a Streamlit app (URL provided).
Elements
• Annotations for protocols (HTTP/REST, SDK calls), formats (image, JSON, PDF).