A comprehensive AI-powered Sales Development Representative (SDR) system built with multi-agent architecture for automated lead generation, research, proposal generation, and outreach.

SalesShortcut is a sales automation and engagement platform that finds, creates, and converts leads through intelligent AI agents. The system automatically discovers potential business leads, researches their needs, creates personalized proposals, and manages outreach campaigns including phone calls and email communication.

SalesShortcut consists of 5 specialized microservices working together:

• 🔍 Lead Finder - Discovers potential business leads in specified cities using Google Maps and location-based search
• 🧠 SDR Agent - Main orchestrator conducting research, proposal generation, and outreach (includes phone calls and email)
• 📋 Lead Manager - Manages lead data, tracks conversion status, and handles meeting scheduling
• 🖥️ UI Client - Web dashboard for monitoring and controlling the entire system
• 📧 Gmail PubSub Service - Handles incoming email responses and lead engagement tracking

graph TD
    A[Lead Finder] --> B[SDR Agent]
    B --> C[Research Lead Agent]
    C --> D[Proposal Generator]
    D --> E[Draft Writer Agent]
    E --> F[Fact Checker Agent]
    F --> G[Outreach Caller Agent]
    G --> H[Lead Clerk Agent]
    B --> I[Outreach Agent]
    I --> J[Phone Call Tool]
    I --> K[Email Tool]
    H --> L[Lead Manager]
    L --> M[Calendar Assistant]
    N[Gmail PubSub] --> L

• Geographic targeting - Find businesses in any specified city
• Google Maps integration - Leverage Maps API for accurate business data
• Intelligent clustering - Group similar businesses for targeted campaigns

• Comprehensive business analysis - Understand target business needs and pain points
• Competitor research - Analyze market position and opportunities
• Review analysis - Extract insights from customer feedback
• Website assessment - Identify digital presence gaps

• Custom proposal generation - AI-crafted proposals based on research findings
• Multi-channel communication - Phone calls via ElevenLabs + email outreach
• Professional voice calls - Natural AI voice conversations with prospects
• Email automation - SMTP-based email delivery with personalized content

• Automated follow-up - Track engagement and schedule next actions
• Status tracking - Monitor leads through the entire sales funnel
• Meeting scheduling - Integrate with calendar systems for appointments
• Data persistence - Store all interactions in BigQuery for analytics

• Framework: Python with FastAPI and Google ADK Agents
• AI Models: Gemini 2.0 Flash Lite (configurable)
• Voice: ElevenLabs API for natural phone conversations
• Database: Google BigQuery for data storage and analytics
• APIs: Google Maps, Google Search, SMTP for email
• Frontend: HTML/CSS/JavaScript with WebSocket real-time updates
• Deployment: Docker containers with Cloud Run support

• Python 3.9+
• Google Cloud Project with BigQuery enabled
• API Keys:
  • Google API Key (for Gemini LLM)
  • Google Maps API Key
  • ElevenLabs API Key (for voice calls)
  • Optional: OpenAI, Anthropic keys

git clone <repository-url>
cd SalesShortcut
cp config.template .env
# Edit .env with your API keys

Update .env with your API credentials:

# Required
GOOGLE_API_KEY=your_google_api_key_here
GOOGLE_MAPS_API_KEY=your_maps_api_key_here
GOOGLE_CLOUD_PROJECT=your_gcp_project_id_here

# For Voice Calls
ELEVENLABS_API_KEY=your_elevenlabs_key_here
ELEVENLABS_AGENT_ID=your_agent_id_here
ELEVENLABS_PHONE_NUMBER_ID=your_phone_number_id_here

# Email Configuration
EMAIL_USERNAME=[email protected]
EMAIL_PASSWORD=your_app_password_here
FROM_EMAIL=[email protected]

# Optional
OPENAI_API_KEY=your_openai_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here

# Install all dependencies
pip install -r requirements.txt

# Or install individual service dependencies
pip install -r lead_finder/requirements.txt
pip install -r lead_manager/requirements.txt
pip install -r sdr/requirements.txt
pip install -r ui_client/requirements.txt
pip install -r gmail_pubsub_listener/requirements.txt

# Start all services
./deploy_local.sh

# Or start individual services
python -m lead_finder --port 8081
python -m lead_manager --port 8082
python -m sdr --port 8084  
python -m ui_client --port 8000
python -m gmail_pubsub_listener --port 8083

# Build and deploy with Docker
make build-all
make deploy-local

# Deploy to Google Cloud Run
./deploy_cloud_run.sh

Open your browser to http://localhost:8000 to access the web interface.

Each service has its own detailed README with specific installation and configuration instructions:

• 🔍 Lead Finder - Find potential leads using Google Maps
• 🧠 SDR Agent - Research and outreach automation
• 📋 Lead Manager - Lead qualification and meeting scheduling
• 🖥️ UI Client - Web dashboard and monitoring
• 📧 Gmail PubSub Service - Email response handling

Each service can have its own .env file for service-specific settings:

• lead_finder/.env - Lead finder specific settings
• lead_manager/.env - Lead manager specific settings
• sdr/.env - SDR agent specific settings
• gmail_pubsub_listener/.env - Gmail service specific settings

Each service has its own Dockerfile and can be deployed independently:

# Build specific service
docker build -f Dockerfile.lead_finder -t salesshortcut-lead-finder .
docker build -f Dockerfile.sdr -t salesshortcut-sdr .
docker build -f Dockerfile.lead_manager -t salesshortcut-lead-manager .
docker build -f Dockerfile.ui_client -t salesshortcut-ui-client .

# Deploy to Cloud Run
./deploy_cloud_run.sh

The system uses Google BigQuery for persistent storage:

• Lead Data: Business information, contact details, research results
• Interaction History: Call transcripts, email responses, follow-up activities
• Analytics: Conversion metrics, campaign performance data

• All API keys stored securely in environment variables
• Phone calls comply with telemarketing regulations
• Business data encrypted in transit and at rest
• No sensitive information logged or exposed

1. Enter Target City: Use the web interface to specify a city for lead generation
2. Automatic Discovery: The system finds businesses without websites in that area
3. View Results: Monitor discovered leads in the dashboard

1. Select Leads: Choose businesses from the lead finder results
2. Automated Research: SDR agent researches each business thoroughly
3. Proposal Generation: AI creates personalized website development proposals
4. Phone Outreach: System makes professional calls to business owners
5. Email Follow-up: Sends detailed proposals via email to interested prospects
6. Lead Tracking: Monitors responses and schedules follow-up activities

• Real-time Updates: WebSocket-powered live status updates
• Lead Management: View, filter, and manage all prospects
• Call Monitoring: Track phone call outcomes and transcripts
• Analytics: Monitor conversion rates and campaign performance

# In any service's config.py
MODEL = "gemini-2.0-flash-lite"  # or "gpt-4", "claude-3-sonnet"
TEMPERATURE = 0.2
TOP_P = 0.95
TOP_K = 40

Services communicate via HTTP/A2A protocol. Default ports:

• UI Client: 8000
• Lead Finder: 8081
• Lead Manager: 8082
• Gmail PubSub: 8083
• SDR: 8084

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

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

For issues, questions, or feature requests:

1. Check the troubleshooting section in individual service READMEs
2. Review service logs for error details
3. Open an issue on GitHub with detailed information

Built with ❤️ for automating the sales development process