ScrumAI - Intelligent Meeting Assistant

AI-powered meeting productivity platform with real-time transcription and intelligent chat assistance

Built for Qualcomm Hackathon NYU 2025

Overview

ScrumAI transforms meeting productivity by combining real-time speech recognition with conversational AI assistance. The platform integrates OpenAI Whisper for precise transcription with AnythingLLM for contextual chat interactions, creating a comprehensive meeting intelligence solution.

Core Features

Real-Time Meeting Intelligence

• Live Audio Transcription: OpenAI Whisper integration for industry-leading accuracy
• Instant Keyword Extraction: AI-powered identification of key discussion points
• Audio Visualization: Dynamic waveform display and level indicators
• Meeting Timer: Automatic session tracking and duration monitoring

AI Chat Integration (NEW)

• AnythingLLM Assistant: Conversational AI with meeting context awareness
• RAG-Powered Queries: Ask complex questions about your meeting content
• Live Context Access: Chat assistant has real-time access to transcript data
• Intelligent Query Routing: Automatic detection of complex vs simple queries
• Streaming Responses: Real-time AI response generation
• Document Indexing: Automatic transcript upload for enhanced retrieval

Export Capabilities

• Notion Integration: Direct export of meeting notes and summaries
• GitHub Integration: Automatic issue creation from action items
• Local Storage: Timestamped transcript files
• Multi-Format Output: Text, JSON, and structured data exports

Architecture

User Interface (Electron Renderer)
    ↓
Main Process (Node.js + IPC)
    ↓
AI Processing Layer (Python)
    ├── Whisper Engine (Real-time STT)
    ├── AnythingLLM (RAG System + Chat API)
    └── Meeting Analysis (NLP Processing)
    ↓
Data Storage (Local Files + External APIs)

Technology Stack

Frontend: Electron 27.0.0, Vanilla JavaScript, Web Audio API
AI/ML: OpenAI Whisper (ONNX), AnythingLLM, PyTorch, Transformers
Backend: Node.js, Python 3.10+, IPC Communication

Quick Start

Prerequisites

• Node.js 18+ (LTS recommended)
• Python 3.10+ with pip
• Git for version control
• Microphone for audio capture

Installation

1. Clone the repository

   git clone https://github.com/anshul-kumar1/scrumAI.git
   cd scrumAI

2. Install dependencies

   npm install
   npm run setup-python

3. Configure environment variables

   cp config.env.example config.env
   # Edit config.env with your API keys

4. Start the application

   npm start

Development Mode

npm run dev  # Starts with DevTools open

Configuration

Environment Variables

Create a config.env file in the project root:

# Notion API Configuration
NOTION_API_KEY=your_notion_api_key_here
NOTION_PARENT_PAGE_ID=your_notion_parent_page_id_here

# GitHub Configuration
GITHUB_TOKEN=your_github_token_here
GITHUB_OWNER=your_github_username_or_org
GITHUB_REPO=your_repository_name

AnythingLLM Setup

Configure AnythingLLM integration in whisper/anythingLLM/config.yaml:

# AnythingLLM Configuration
api_key: "your_anythingllm_api_key"
model_server_base_url: "http://localhost:3001/api"
workspace_slug: "your_workspace_name"
stream: true
stream_timeout: 30

API Setup

Notion Integration

1. Create a Notion integration
2. Copy the API key to NOTION_API_KEY
3. Share a page with your integration and copy the page ID

GitHub Integration

1. Create a GitHub personal access token
2. Grant repo permissions for issue creation

AnythingLLM Setup

1. Install and run AnythingLLM
2. Create a workspace and obtain API key
3. Update the config file with your local AnythingLLM instance details

Usage Guide

Starting a Meeting

1. Launch ScrumAI and grant microphone permissions
2. Click the "Start Meeting" button
3. Begin speaking - real-time transcription starts immediately
4. Monitor live transcript, keywords, and audio visualization
5. Use the Chat tab to ask questions about the meeting

Using the AI Chat Assistant

Simple Queries (Live Context)

• "What did we just discuss about the authentication system?"
• "Who is taking lead on the API development?"
• "What was the last action item mentioned?"

Complex Analysis (RAG-Powered)

• "Summarize the key decisions made in this meeting"
• "What are the main technical challenges identified?"
• "Create a list of all action items with assigned owners"
• "Analyze the sentiment of the discussion about the Q2 release"

The system automatically determines whether to use live context or RAG based on query complexity.

Post-Meeting Workflow

1. Review content using transcript and keywords tabs
2. Ask the chat assistant for summaries and insights
3. Edit and refine transcript or AI-generated content
4. Export to Notion for documentation or GitHub for issue tracking
5. Automatic local backup with timestamp

Build & Distribution

npm run build         # Development build
npm run build:win     # Windows (NSIS installer)
npm run build:mac     # macOS (DMG)
npm run build:linux   # Linux (AppImage)

Project Structure

scrumAI/
├── src/
│   ├── electron/              # Main Electron process
│   ├── renderer/              # Frontend application
│   │   ├── js/                # JavaScript modules
│   │   │   ├── chat-controller.js  # AnythingLLM chat interface
│   │   │   └── audio-manager.js    # Web Audio API integration
│   │   └── index.html         # Main UI structure
│   └── services/              # Backend services
│       ├── whisperService.js  # Whisper transcription service
│       └── chatbotService.js  # AnythingLLM chatbot service
├── whisper/                   # AI processing layer
│   ├── anythingLLM/          # AnythingLLM integration
│   │   ├── chatbot_client.py  # Python AnythingLLM client
│   │   └── config.yaml        # LLM configuration
│   ├── meeting_transcriber.py # Enhanced transcriber with LLM
│   └── models/               # ONNX model files
└── config.env               # Environment configuration

Troubleshooting

Audio not working:

• Verify microphone permissions in system settings
• Ensure no other applications are using the microphone

Python AI processing fails:

• Confirm Python 3.10+ installation: python --version
• Reinstall dependencies: npm run setup-python

AnythingLLM chat not responding:

• Verify AnythingLLM service is running on localhost:3001
• Check API key configuration in whisper/anythingLLM/config.yaml
• Ensure workspace exists and is accessible

Export features not working:

• Validate API keys in config.env
• Test internet connectivity
• Verify service permissions (Notion, GitHub)

Testing

npm test        # Run test suite
npm run lint    # Code quality check

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Qualcomm for the hackathon opportunity
• NYU for hosting the event
• OpenAI Whisper for speech recognition capabilities
• AnythingLLM for conversational AI platform
• Notion and GitHub for integration APIs

---

Built by the ScrumAI Team for Qualcomm Hackathon NYU 2025