---

Meeting Assistant Web Application

Overview

The Meeting Assistant is a Flask-based web application that allows users to upload meeting audio recordings, transcribe them into text, and extract actionable items such as emails, to-dos, calendar events, and notes. The system features secure authentication and integrates with external APIs like Groq for transcription and NLP-based action item extraction.

The WebApp is deployed in this link = Link

---

Features

• User Authentication: Secure login/logout system with token-based session management.
• Audio Transcription: Uses Groq Whisper API to transcribe audio files.
• Action Item Extraction: NLP-powered classification of tasks like emails, notes, calendar events, etc.
• Action Item Management: Accept, reject, update, and view categorized action items.
• Frontend Integration: Basic HTML templates for login and home/dashboard.

---

Table of Contents

1. Installation
2. Project Structure
3. Configuration
4. Authentication
5. API Endpoints
6. Action Items
7. Audio Transcription
8. Frontend Integration
9. Error Handling
10. Security Considerations
11. Performance & Scalability
12. Potential Future Enhancements
13. Troubleshooting
14. Contributing
15. License
16. Acknowledgements

---

Installation

1. Clone the Repository

git clone <repository_url>
cd <project_directory>

2. Create and Activate a Virtual Environment

python -m venv venv
source venv/bin/activate  # Linux/MacOS
venv\Scripts\activate     # Windows

3. Install Dependencies

pip install -r requirements.txt

4. Set up Environment Variables

Create a .env file in the project root:

FLASK_SECRET_KEY=<your_secret_key>
BASE_URL=<backend_base_url_for_auth>
GROQ_API=<your_groq_api_key>
APP_URL=<your_app_url>

5. Run the Application

python main.py

---

Project Structure

/project_root
├── /static                # Static files (CSS, JS, images)
├── /templates             # HTML templates
│   ├── login.html
│   ├── index.html
├── /temp                  # Temporarily stored audio files
├── /groq                 # Groq API client wrapper
├── /instructor           # NLP classification logic
├── main.py                # Main Flask app
├── .env                   # Environment configuration
├── requirements.txt       # Project dependencies

Key Files:

• main.py: Application logic and routes
• /templates: Login and home page HTML
• /static: Frontend assets
• /temp: Temporary audio storage

---

Configuration

Environment Variables:

• FLASK_SECRET_KEY: Secret key for sessions
• BASE_URL: Auth backend base URL
• GROQ_API: Groq API key
• APP_URL: App base URL

File Paths:

• /static – CSS/JS
• /templates – Frontend HTML
• /temp – Audio file uploads

---

Authentication

• Login Route (POST /): Authenticates with email and password
• Session Management: Flask sessions store tokens securely
• Home Route (GET /home): Protected route showing user dashboard

---

API Endpoints

Authentication

• POST /
  Authenticates user and redirects to home on success.

• GET /home
  Loads the authenticated user's dashboard.

---

Transcription

• POST /transcribe
  Uploads audio and returns raw transcription text.

• POST /transcribe-and-extract
  Uploads audio and returns structured action items.

---

Action Item Processing

• POST /extract-action-items
  Accepts transcript and returns action items.

• POST /update-json-file
  Updates a specific item in the JSON store.

• POST /reject-action-item
  Deletes a rejected item.

• POST /accept-action-item
  Sends accepted items to the appropriate service (e.g., email, calendar).

---

Action Items

Categories:

• Email – Tasks requiring email communication
• Web Search – Research or lookup tasks
• Note – Notes or reminders
• To-Do – Tasks to complete
• Calendar Event – Scheduling items

Action items are saved in categorized JSON files (emails.json, todos.json, etc.).

Extraction is powered by Groq + Instructor API.

---

Audio Transcription

• Audio files are uploaded and temporarily stored in /temp.
• Groq Whisper API performs transcription.
• Resulting text is returned or passed for action item extraction.

---

Frontend Integration

Login Page (login.html)

• Simple form for email/password
• Stores token in browser on success

Home Page (index.html)

• Dashboard UI
• Shows extracted items and action controls

---

Error Handling

Handles:

• Missing or invalid files/fields
• Auth failures
• External API errors
• Invalid action item indexes or types

---

Security Considerations

• Session Tokens: Securely stored in cookies
• Password Hashing: (Recommend using bcrypt or werkzeug.security)
• Token Expiry: (JWT or session expiry is recommended)
• File Validation: Accept only audio/* MIME types
• API Key Management: Environment-based access control

---

Performance & Scalability

• Large File Uploads: Consider Flask upload size limits
• Asynchronous Processing: Offload transcription to Celery for better performance
• Caching: Use Redis to cache repeated transcripts
• Storage: Integrate cloud storage (e.g., S3) for large-scale deployments
• Database: Migrate from JSON to PostgreSQL/MySQL in production

---

Troubleshooting

• Missing .env: Ensure all variables are set
• Invalid audio format: Accept only audio/* types
• Rate limits: Handle Groq API limits with retries
• Large files: Use Celery for background tasks

---

Contributing

We welcome contributions!

1. Fork this repo
2. Make changes (code, docs, fixes)
3. Submit a pull request

Code Style:

• Python: PEP8
• JS: ESLint
• Document your changes!

---

License

This project is licensed under the MIT License. See LICENSE for details.

---

Acknowledgements

• Groq API – Transcription services
• Flask – Web framework
• Whisper – Speech-to-text model
• Instructor – NLP-based action item extractor

---