AI Meeting Assistant is an intelligent post-meeting automation platform built to transform how teams manage and revisit meetings. By leveraging advanced AI and a modular architecture powered by OPEA microservices, this tool eliminates the need for manual note-taking, enhances productivity, and makes meetings more accessible and actionable.

Users can effortlessly upload their meeting data—whether transcripts or audio/video recordings (e.g., .mp3, .wav, .mp4)—and let the assistant handle the rest. The system processes content automatically, providing valuable insights, summaries, and actionable outcomes without having to rewatch or re-read lengthy discussions.

• 🎙 Automatic Transcription
  Converts audio/video recordings into accurate, searchable text.

• 📝 Meeting Summarization
  Generates concise summaries for quick review and better understanding.

• 📋 Actionable Minutes of Meeting (MoM)
  Automatically detects and organizes key points, decisions, and action items from the discussion.

• ❓ Semantic Q&A
  Enables users to query the meeting content using natural language questions and receive context-aware answers.

• 😊 Sentiment Analysis
  Performs line-by-line speaker sentiment analysis to gauge tone and engagement.

• Save time by eliminating manual note-taking
• Quickly identify key decisions and follow-ups
• Enhance team collaboration and accountability
• Revisit meeting content effortlessly using search and Q&A
• Improve inclusivity with multilingual support and transcription accessibility

• Framework: FastAPI (Python)
• Database: PostgreSQL (via SQLAlchemy ORM)
• Authentication: JWT-based authentication with OTP verification
• File Storage: Local server file system
• Vector Database: Redis with RediSearch

• ASR Microservice: Converts audio/video files to text using automatic speech recognition.
• Dataprep Microservice: Cleans and processes transcript text and stores embeddings in the vector database.
• Embeddings Microservice: Generates vector embeddings from textual data.
• Retriever Microservice: Enables fast semantic search across meeting content.
• LLM Microservice: Powers natural language Q&A and generates contextual summaries.
• Document Summary Microservice: Extracts concise Minutes of Meeting (MoM) from transcripts.

• Build Tool: Vite
• Framework: React.js
• UI Library: Material UI
• State Management: Redux Toolkit
• Repo Link: https://github.com/sam-79/SyncAgenda

• Docker: Used for microservice containerization and deployment

Follow these steps to set up and run the AI Meeting Assistant locally.

git https://github.com/shantanu1905/meeting-ai-assistant.git
cd ai-meeting-assistant

Create a .env file in both the backend directories (or wherever appropriate), and add necessary environment variables.

# Retrieve environment variables
POSTGRES_HOST=127.0.0.1
POSTGRES_DB=postgres
POSTGRES_USER=postgres
POSTGRES_PASSWORD=postgres
POSTGRES_PORT=5432


HOST_URL = 127.0.0.1:8000

JWT_SECRET=e56623570e0a0152989fd38e13da9cd6eb7031e4e039e939ba845167ee59b496
RABBITMQ_URL=localhost


GMAIL_ADDRESS=your-mail
GMAIL_PASSWORD=your-mail-password
HUGGING_FACE_API_KEY=your-api-key

REDIS_HOST=localhost
REDIS_PORT=6379
REDIS_DB=0

Make sure you have Docker and Docker Compose installed.

docker-compose up 

This will:

Start all OPEA microservices

Launch PostgreSQL and Redis.

pip install -r requirements.txt
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

API Docs available at: http://localhost:8000/docs