Why Your AI Has No Idea What's Happening in the World with Campbell Brown, CEO at PredictHQ

Description

Understanding the real world is the biggest potential advantage of AI. Today, we're talking to Campbell Brown, CEO and cofounder of PredictHQ, about the layer of real-world intelligence that AI models are still missing. We discuss why 65% of unexplained demand spikes come down to something most businesses completely ignore, how a New Zealand car rental company became the data backbone for Uber and Expedia, why the era of reading API documentation is effectively over, and what it actually means to "eat stress" as a founder — and why reconciling it matters just as much. All of this right here, right now, on the Modern CTO Podcast!  To learn more about PredictHQ, check out their website here: https://www.predicthq.com/ 00:00 – Introduction: What PredictHQ Does 01:48 – Campbell's Background and Moving to America 02:18 – How Uber Became Their First Customer 03:19 – Pivoting from SaaS to API-First 04:39 – The Car Rental Origin Story 05:52 – Managing 10,000 Data Sources at Scale 07:47 – Event Planning Companies and Wallet Share 09:59 – Company Size and Profitability 11:42 – The Small Team, High Revenue Per Employee Shift 13:48 – Why Everyone Tries to Build This Internally 15:17 – Introducing Bolt: PredictHQ's AI Developer Environment 16:33 – How Beam Personalizes Predictions by Business 18:49 – The MCP Integration with Claude and ChatGPT 20:39 – Why AI Hallucinations Compound at Scale 21:31 – Domino's, Expedia, and the Largest US Retailer 25:39 – Bolt vs. MCP: What Each Tool Actually Does 32:02 – Campbell's Founder Journey and Early Career 33:47 – Leadership Advice: Making Your Bed 35:32 – Eating Stress and the Importance of Reconciling It 37:05 – Shared Founder Experiences: Payroll Nights and Thick Skin 39:20 – Pattern Recognition as a Founder Superpower 40:25 – Sign-Off