Built for a B2B software company selling AI-powered dispatch and fleet-control tools into the DACH logistics market (Germany, Austria, Switzerland). Every lead is automatically researched via LinkedIn and company-website scraping, then handed to Claude to write a genuinely personalised cold email — not a mail-merge template with blanks filled in. Delivery is spread across 15 rotating inboxes with capped daily limits and human-like send timing to protect deliverability, and every open, reply, and follow-up is tracked and reported automatically.
What it does
- Automated lead research — pulls LinkedIn profile data and scrapes the company website for real context, with fallback extraction so no lead silently drops out of the pipeline
- AI-written, not templated — Claude writes each email from a fixed brand voice and structure, grounded in that specific lead’s research
- Multi-inbox sending — rotates delivery across 15 separate inboxes with capped daily volume and randomised human-like delays to protect sender reputation
- Full reply and open tracking — replies are pulled and cleaned automatically from every inbox, opens tracked in real time, follow-ups sent automatically to non-responders
- Self-correcting — missing data, failed AI generation, or failed sends are caught, logged, and pushed to a real-time ops alert channel instead of silently failing
4,137
Emails sent (30 days, autonomous)
690+
Hours of manual work eliminated
Built with n8n · Claude (Anthropic) · Gmail API · Google Sheets · Custom live dashboard · Oxylabs Studio · BrightData
Read full case study →