An open-source AI GTM system that replaces a $150K hire with ~$100/month in tools.

Built with OpenClaw — the AI agent framework.

LinkedIn Engagement → Verified Emails → Personalized Outreach → Booked Meetings

This kit automates the entire top-of-funnel GTM process:

1. Mine LinkedIn — Find people engaging with niche content (warm leads, not cold lists)
2. Enrich — Get verified work emails via Hunter.io + Apollo
3. Score — Rank leads against your ICP (A/B/C/D tiers)
4. Write — Generate hyper-personalized cold emails using AI
5. Send — Load into Instantly.ai sequences

The result: A full pipeline from "who's talking about my space?" to "meeting booked" — running on autopilot.

I've spent 20 years in marketing. Scaled Fireball Whisky from one state to a billion-dollar global brand. Ran campaigns for Heineken, Hennessy, Buffalo Trace. Now I run Emerald Digital, an AI-first marketing agency.

Here's what I learned: Most startups can't afford a GTM engineer. So founders do outbound manually (slow), hire too early (expensive), or skip it entirely (fatal).

This kit is the system I wish I had. It does what a $150K/year GTM hire does — for about $100/month in API costs.

I'm open-sourcing it because the best marketing happens when founders can actually reach their customers.

git clone https://github.com/themattberman/first-1000-kit.git
cd first-1000-kit

cp .env.example .env
cp brand-config.example.json brand-config.json

# Edit both files with your keys and ICP

# Full pipeline: find 50 leads in your niche
./run.sh "AI marketing automation" 50

# Or run with OpenClaw agent
openclaw start
# Then message: "Mine LinkedIn for AI marketing, 50 leads"

See SETUP.md for detailed instructions.

Each skill can run standalone or as part of the full pipeline.

┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│  LinkedIn       │────▶│   Enrichment    │────▶│   ICP Scoring   │
│  Mining         │     │   (emails)      │     │   (A/B/C/D)     │
└─────────────────┘     └─────────────────┘     └─────────────────┘
                                                        │
                                                        ▼
┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│  Instantly.ai   │◀────│   Outreach      │◀────│   A-Tier        │
│  Sequences      │     │   Writer        │     │   Leads         │
└─────────────────┘     └─────────────────┘     └─────────────────┘

Traditional outbound scrapes LinkedIn for titles and blasts generic emails. Response rates: 1-2%.

This system finds people actively engaging with content in your niche. They commented on a post about AI marketing? They're already thinking about it. That's a warm lead.

Response rates with warm + personalized: 8-15%.

Edit brand-config.json to define:

{
  "icp": {
    "target_titles": ["VP Marketing", "Head of Growth", "CMO"],
    "target_industries": ["SaaS", "E-commerce"],
    "company_size": { "min": 50, "max": 500 }
  },
  "outreach": {
    "sender_name": "Your Name",
    "tone": "professional but conversational",
    "pain_points": ["Problem 1 you solve", "Problem 2"]
  }
}

Or just tell the OpenClaw agent your ICP conversationally — it'll figure it out.

vs hiring a GTM engineer: $150,000/year

This kit is built for OpenClaw, an open-source AI agent framework.

# Install OpenClaw
npm install -g openclaw

# Copy kit to workers directory
cp -r first-1000-kit ~/clawd/workers/first-1000

# Start the agent
cd ~/clawd/workers/first-1000
openclaw start

Then just message it naturally:

• "Mine LinkedIn for people talking about AI automation"
• "Enrich and score against my ICP"
• "Write outreach for A-tier leads"
• "Load to Instantly"

The agent handles the orchestration.

first-1000-kit/
├── README.md              # You're here
├── SETUP.md               # Detailed setup guide
├── run.sh                 # Pipeline runner
├── .env.example           # API key template
├── brand-config.example.json  # ICP template
├── skills/
│   ├── linkedin-miner/
│   ├── lead-enricher/
│   ├── email-verifier/
│   ├── icp-scorer/
│   ├── outreach-writer/
│   ├── instantly-loader/
│   └── pre-call-research/
├── SOUL.md                # Agent personality (for OpenClaw)
├── AGENTS.md              # Agent instructions
└── SPEC.md                # Full system spec

This is open source. PRs welcome.

Ideas for contribution:

• Additional lead sources (Twitter/X, Reddit, etc.)
• More email providers (Lemlist, Apollo sequences)
• Better scoring models
• UI/dashboard

MIT License. Use it, fork it, build on it.

Built by Matt Berman.

• 🐦 Twitter/X: @themattberman
• 📰 Newsletter: Big Players
• 🏢 Agency: Emerald Digital

This is how bootstrapped founders compete with funded teams.

Star the repo if this helps. It tells me to keep building.