Client project — automated B2B lead generation and personalized outreach
Who this is for
B2B wholesalers and distributors who need a steady pipeline of qualified leads — but don't have time to manually research hundreds of companies and write personalized outreach for each one.
The problem
Finding good B2B leads means researching each company's website, industry, and size, then writing a personalized email or LinkedIn message for every single one. Doing this by hand for hundreds of prospects is slow, repetitive, and most outreach ends up generic — which gets ignored.
What I built
An end-to-end AI lead generation and outreach pipeline:
- Multi-channel AI agent (WhatsApp-style webhook) that reads voice notes, images, and text — handles buyer replies, meeting requests, and order confirmations automatically, with GPT-4o Vision reading product images and Whisper transcribing voice messages
- Automatically checks calendar availability and books meetings via Google Meet when a buyer wants to talk
- A lead research pipeline that pulls company details (industry, website, address, phone) into Supabase
- GPT-4o scores each lead's fit (1–10) and readiness (cold / warm / hot) based on company signals
- Generates a unique, personalized email subject, email body, and LinkedIn DM for every single lead — referencing their actual business, not a generic template
Tech stack
n8n, Supabase, OpenAI GPT-4o (+ Vision), Whisper, Google Calendar/Meet API, Gmail API
The result
- 187 leads researched, scored, and given ready-to-send personalized outreach automatically
- Every lead comes with a fit score and readiness label, so the client knows exactly who to contact first
- No generic templates — every email and LinkedIn message is written specifically for that company
- Buyer replies (text, voice, or image) are understood and responded to automatically, with meetings booked without any manual back-and-forth
