AI-powered lead filter. Accepts data from any source, uses AI to research and qualify leads, then displays them in a dashboard and sends digest emails.

1. Ingest leads - POST to the endpoint with any data:

   POST https://valdottown-leads.web.val.run?source=rb2b
   Body: { "email": "lead@example.com", "company": "Acme" }
   

2. View dashboard - Visit the endpoint URL in your browser

3. Configure digest - Edit digest.ts to add email recipients

1. Data arrives via POST with ?source=your-source
2. AI researches the person and extracts: name, company, role, email, LinkedIn
3. AI qualifies them against your ICP: yes, no, or insufficient_data
4. Lead is stored and visible in dashboard
5. Daily digest emails ideal leads

Just point your webhook to: https://your-val.web.val.run?source=source-name

Examples:

• RB2B: ?source=rb2b
• Tally form: ?source=tally-form
• Zapier: ?source=zapier

Create a new interval file that calls processLead:

Create val
import { processLead } from "./main.ts";

export default async function cron() {
  const newData = await fetchFromSomewhere();
  for (const item of newData) {
    await processLead(item, "my-source");
  }
}

Edit PROMPT.txt to describe your ideal customer profile.

Edit digest.ts and add emails to the RECIPIENTS array.

• main.ts - HTTP endpoint + routing
• db.ts - Database schema + queries
• dashboard.tsx - Web UI
• digest.ts - Email digest cron
• agent.ts - AI qualification logic
• PROMPT.txt - ICP prompt for AI
• testing.ts - Test utilities