CalPilot

An AI-powered call analysis platform for reviewing and analyzing sales calls.

Developed by Adeel Zafar

Local Development Setup

1. Copy example environment file and configure your keys:

cp .env.example .env

2. Create and activate virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Set up local environment:

# Copy example environment file
cp .env.example .env

# Edit .env.local with your settings
nano .env

# Create development data directory
mkdir -p data
cp your_calls.json data/calls.json

5. Run the application:

flask run

Project Structure

• app/: Main application package
  • api/: API endpoints
  • services/: Business logic
  • templates/: HTML templates
  • static/: Static files
• data/: Local development data
• config.py: Configuration settings
• tests/: Test suite

Development

• Built with Flask and Anthropic's Claude API <3
• Frontend using Tailwind CSS
• Comprehensive test suite
• Caching for performance
• Environment-specific configurations

Core Technologies

• Flask: Lightweight web framework with excellent extensibility
• Anthropic Claude API: State-of-the-art AI for natural language processing
• Google Cloud Platform: Scalable cloud infrastructure
  • Cloud Run for containerized deployment
  • Cloud Storage for secure data management

Key Libraries

• anthropic: Official Claude AI SDK for intelligent call analysis
• python-dotenv: Secure environment configuration management
• gunicorn: Production-grade WSGI server
• pytest: Comprehensive testing framework

Advanced Features

• Intelligent Caching:

  • LRU cache for optimized duration formatting
  • Time-based cache invalidation for call data
  • Environment-specific cache strategies

• Robust Logging:

  • Rotating file handlers with 1MB limit
  • Structured logging with line numbers
  • Environment-aware log levels
  • Comprehensive error tracking

• Error Handling:

  • Graceful API error management
  • Rate limiting protection
  • Timeout handling
  • Detailed error reporting

• Security:

  • Environment-based configurations
  • Secure credential management
  • Cloud-native security practices

• Testing:

  • using pytest
  • 

Features

Recent Calls

Filter calls by date

Search for calls

List of companies

Call summary and Key topics

Pre-determined questions for each call

Asking your own questions

Answer from Claude <3

License

MIT License