README.md

metadata

title: Handwritten Notes To Summary
emoji: 🦀
colorFrom: blue
colorTo: red
sdk: gradio
sdk_version: 6.5.1
app_file: app.py
pinned: false
license: mit
short_description: 'Extracts Handwritten notes into text and summary '

Whiteboard Notes → Meeting Summary

Upload photos of whiteboard or handwritten meeting notes and get a clean, structured summary with action items.

What This Does

You take photos of a whiteboard after a meeting. The app reads all the text, even messy handwriting, and turns it into organized meeting notes. It pulls out the key decisions, action items, who is responsible, and due dates. You can copy the result to Slack or Notion, or download it as a Word document.

Live Demo

Try it here: [Your HuggingFace Space URL]

Features

• Upload one or multiple whiteboard photos
• AI powered text extraction that handles messy handwriting
• Automatic organization into sections: Summary, Decisions, Action Items
• Action items table with Owner, Due Date, and Priority columns
• Highlights items that need clarification (missing owners, dates, etc.)
• Download as Word document
• Copy paste ready format for Slack and Notion

How It Works

1. You upload photos of your whiteboard or handwritten notes
2. The AI vision model extracts all visible text using OCR
3. The text is analyzed and structured into meeting notes format
4. Action items are identified and organized into a table
5. Missing information (like owners or dates) is flagged for follow up
6. You get a clean summary ready to share

Quick Start

1. Go to the app
2. Upload one or more photos of your whiteboard notes
3. Optionally add meeting context like "Weekly standup" or "Project kickoff"
4. Click "Process Notes"
5. Wait about 30 seconds for processing
6. Copy the summary or download the Word document

Running It Yourself

What You Need

• Python 3.10 or higher
• A HuggingFace account (free)
• A HuggingFace API token

Setup

Clone the repo:

git clone https://github.com/yourusername/whiteboard-to-summary
cd whiteboard-to-summary

Install requirements:

pip install -r requirements.txt

Set your HuggingFace token:

export HF_TOKEN=your_token_here

Run the app:

python app.py

Open your browser to http://localhost:7860

Deploying to HuggingFace Spaces

1. Create a new Space on HuggingFace and select Gradio as the SDK
2. Upload all the files from this repo
3. Go to Settings and add a secret called HF_TOKEN with your token
4. The app will build and start automatically

File Structure

whiteboard-to-summary/
├── app.py              Main application
├── requirements.txt    Python packages needed
└── README.md           This file

Output Format

The app generates meeting notes with these sections:

Meeting Summary - Key points from the discussion in bullet form

Key Decisions - Any decisions that were made during the meeting

Action Items - A table with columns for:

• Action Item (the task)
• Owner (who is responsible, or TBD if not specified)
• Due Date (when it's due, or TBD if not specified)
• Priority (High, Medium, or Low based on context)

Items Needing Clarification - Things that were unclear or need follow up

Raw Notes - Cleaned up version of the original text for reference

Tips for Good Results

• Make sure the whiteboard is well lit when taking photos
• Keep the camera steady so text is in focus
• Include the entire whiteboard in the frame
• If there are multiple boards, upload all photos at once
• Add meeting context to help the AI understand the notes better

Tech Stack

• Gradio for the web interface
• HuggingFace Hub for AI model access
• Qwen2.5 VL for vision and text understanding
• python-docx for Word document generation

Limitations

• Works best with English text
• Very messy handwriting may not be fully readable
• The AI makes its best guess when text is unclear
• Always review action items and owners before sharing

License

MIT License. Use it however you want.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference