A letter to the Profound team.
ProfoundLoops was built in ~9 hours at Profound's inaugural Marketing Engineer Hackathon. Here's the launch film — and a thank-you.
ProfoundLoops — the earned-AI-visibility engine · 1:18
A deeply oversubscribed room · ProfoundHQ, NYC
Dear Profound team,
Saturday at Profound did not disappoint. I came up just short in your inaugural “Marketing Engineer” Hackathon — but I'm stoked to have been a finalist among ~50 builders in a deeply oversubscribed room at ProfoundHQ in NYC. Congratulations to Aaron Lu and Brianna on the win.
This was my second hackathon since my time at Carvana, and my first dedicated solo-hack event. The ask: “Find a marketing process that's inhuman in scope or scale, and ship a system or agent that runs it” — in about nine hours. We each got a Profound instance loaded with a large AEO dataset, plus full access to the API, MCP server, and Agent Workflows.
Here's the thesis I built around. AI answers are slowly going pay-to-play — OpenAI, AI Mode on Google. But you can't buy your way into Claude, Gemini, or Perplexity, so earned citation is the only known way in. It's brutal to earn, though. So I built ProfoundLoops: an earned-AI-visibility action engine that runs on top of Profound, with Profound agents. You earn the citations you'd otherwise buy, and close the AEO loop — turning organic inbound into attributable, personalized, targeted outbound: email and paid retargeting.
How ProfoundLoops works:
- Pull live AEO/GEO data from Profound — via the API, over MCP.
- Find the exact gaps where you're missing, and generate content to win them.
- When a buyer engages, reverse-match the visit to a company and surface the buying committee — Enrich Labs + Clay.
- Draft personalized outbound — email via Resend — and targeted ads per segment: Google, LinkedIn, TikTok, Meta.
With more time, I'd have run a live ad campaign and pulled the attribution data back into ProfoundLoops via Composio — to show how AEO can be a rising tide that lifts all ships.
