Call Insights: Live Call Analytics Platform

Impetus × AWS

Call Insights is a full-stack application that provides real-time and post-call analytics for call centers and customer support teams. It leverages AI-powered transcription, sentiment analysis, and data management to enhance call center operations.

Screenshot

Features

• AI Chatbot: Interact with an AI assistant for instant support and insights.
• Voice Transcriber: Real-time speech-to-text transcription using your microphone.
• Call Simulator: Practice and analyze simulated customer interactions with live insights.
• Data Manager: Upload, manage, and analyze call recordings and documents (text, PDF, images, audio).
• Live Analytics: Real-time and historical insights into call data.

Tech Stack

• Frontend: React, Vite, Tailwind CSS, Radix UI, React Router
• Backend: FastAPI, Uvicorn, Boto3 (AWS S3/Transcribe), WebSockets, Python
• Cloud: AWS S3 (storage), AWS Transcribe (speech-to-text)

Project Structure

.
├── backend/
│   ├── main.py              # FastAPI backend with REST & WebSocket endpoints
│   ├── requirements.txt     # Python dependencies
│   └── .env                 # AWS credentials and config
└── frontend/
    ├── src/                 # React source code
    ├── package.json         # Frontend dependencies and scripts
    └── vite.config.js       # Vite config (with API proxy)

Getting Started

Prerequisites

• Node.js (v18+ recommended)
• Python 3.8+
• AWS account with S3 and Transcribe permissions

Backend Setup

1. Install dependencies:

   cd backend
   python3 -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt

2. Configure AWS credentials: Create a .env file in backend/ with:

   AWS_ACCESS_KEY_ID=your-access-key
   AWS_SECRET_ACCESS_KEY=your-secret-key
   S3_BUCKET_NAME=your-s3-bucket
   AWS_DEFAULT_REGION=us-east-1
   

3. Run the backend:

   uvicorn main:app --reload --host 0.0.0.0 --port 8000

Frontend Setup

1. Install dependencies:

   cd frontend
   npm install

2. Run the frontend:

   npm run dev

   The app will be available at http://localhost:5173 (or as shown in your terminal).

3. API Proxy: The frontend is configured (via vite.config.js) to proxy /api requests to the backend at http://localhost:8000.

Usage

• Dashboard: Overview of features and quick navigation.
• AI Chatbot: Chat with the AI assistant (API integration required).
• Voice Transcriber: Click "Start Recording" to transcribe speech in real-time.
• Call Simulator: Select a scenario and simulate a customer call with live insights.
• Data Manager: Upload and manage files; supported formats: .txt, .pdf, .jpg, .png, .wav, .mp3.

Environment Variables

• Backend requires AWS credentials and S3 bucket info in .env.

• Frontend does not require environment variables for local development.

• Deploy the backend (FastAPI) to a cloud server or service (e.g., AWS EC2, Heroku).

• Deploy the frontend (React) to a static host (e.g., Vercel, Netlify) and configure the API endpoint accordingly.

License - MIT