Network Bounty Hunter Agent 🎯

AI-powered LinkedIn connector matchmaking system that monitors LinkedIn posts for "I need X" requests and intelligently matches them with relevant contacts from your network.

Built with Google Cloud Platform Agent Starter Pack (adk_base).

✨ What It Does

This agent transforms your LinkedIn network into a powerful matchmaking engine:

Tell the agent about a LinkedIn post:

"I saw a post: Need sustainable packaging supplier for food products"

The agent will:

1. Extract the need from the post (e.g., sustainable packaging expertise, food industry experience)
2. Match contacts from your network based on skills, industry, and relevance
3. Generate personalized introduction messages tailored to the opportunity
4. Summarize the opportunity with actionable next steps

🚀 Quick Start

Prerequisites

• Python 3.10+
• Google Cloud account with Vertex AI enabled
• gcloud CLI authenticated

Installation

# Clone the repository
git clone https://github.com/chance600/network-bounty-hunter-agent.git
cd network-bounty-hunter-agent

# Install dependencies
pip install -r requirements.txt

# Configure Google Cloud (if not already done)
export GOOGLE_CLOUD_PROJECT="your-project-id"
export GOOGLE_CLOUD_LOCATION="global"
export GOOGLE_GENAI_USE_VERTEXAI="True"

Usage

# Start the agent in chat mode
cd app
python -m google.adk.cli chat agent:app

Example conversation:

You: I saw this LinkedIn post: "Looking for a React developer with 5+ years experience for fintech startup in SF"

Agent: I've analyzed the post and found 3 strong matches from your network:

1. Sarah Chen (Relevance: 9.5/10)
   - Senior React Developer at Stripe
   - 7 years experience
   - Based in San Francisco
   - Skills match: React, TypeScript, Fintech

2. Michael Rodriguez (Relevance: 8.7/10)
   ...

Would you like me to generate an introduction message for Sarah?

You: Yes, generate intro for Sarah

Agent: Here's a personalized introduction:

"Hi [Post Author],

I noticed your search for a React developer with fintech experience. I'd like to introduce you to Sarah Chen, a Senior React Developer at Stripe with over 7 years of experience building production-grade fintech applications..."

🛠️ Architecture

Core Components

• app/agent.py - Main agent with 4 connector tools:

  • extract_need_from_post - Parses LinkedIn posts to identify needs
  • find_matching_contacts - Ranks contacts by relevance
  • generate_intro_message - Creates personalized introductions
  • create_opportunity_summary - Provides actionable summaries

• app/models.py - Data models:

  • Contact - Network contact with skills, industry, location
  • Need - Extracted requirements from posts
  • OpportunitySummary - Matchmaking results and recommendations

Contact Database

The agent uses a mock contact database in app/agent.py. Replace MY_CONTACTS with your actual network data:

MY_CONTACTS = [
    Contact(
        id="c1",
        name="Your Contact Name",
        title="Their Job Title",
        company="Their Company",
        industry="Their Industry",
        location="City, State",
        skills=["Skill 1", "Skill 2", "Skill 3"],
        relationship_strength=9,  # 1-10 scale
        linkedin_url="https://linkedin.com/in/username"
    ),
    # Add more contacts...
]

📋 Use Cases

1. Talent Matchmaking

Post: "Hiring: ML Engineer with NLP experience for healthcare startup" Agent: Matches contacts with ML/NLP skills in healthcare industry

2. Business Development

Post: "Looking for sustainable packaging suppliers for CPG brand" Agent: Finds contacts in packaging/manufacturing with sustainability focus

3. Partnership Opportunities

Post: "Seeking co-founder with ops experience for SaaS startup" Agent: Identifies contacts with operations background and startup experience

4. Consulting/Advisory

Post: "Need expert in FDA regulatory compliance for medical devices" Agent: Matches regulatory affairs professionals in medical device sector

🔧 Customization

Adjust Matching Algorithm

Edit scoring logic in find_matching_contacts() tool:

# Adjust weights for different factors
skill_match_weight = 0.4  # 40% weight on skills
industry_match_weight = 0.3  # 30% weight on industry
relationship_weight = 0.2  # 20% weight on relationship strength
location_weight = 0.1  # 10% weight on location proximity

Custom Introduction Templates

Modify message generation in generate_intro_message() tool to match your style.

🧪 Testing

# Test with sample scenarios
python -m google.adk.cli chat agent:app

# Sample test inputs:
> "Post says: Need React dev in NYC"
> "Show me matches for: sustainable packaging supplier"
> "Generate intro for contact ID c1"

📚 Documentation

See docs/IMPLEMENTATION_PLAN.md for detailed implementation notes and architecture decisions.

🔐 Privacy & Security

• Contact data stays local (no scraping)
• No automated LinkedIn posting
• You maintain full control over introductions
• All API calls use authenticated Google Cloud services

📝 License

MIT License - see LICENSE file

🤝 Contributing

Contributions welcome! Please open an issue or PR.

🙏 Acknowledgments

Built with Google Cloud Platform Agent Starter Pack

---

Ready to turn your network into a matchmaking engine? Star this repo and start connecting! ⭐