#379 – How Zohar Bronfman built AI that actually delivers ROI

Episode notes

A story about rejecting the magic wand approach to AI—and building something businesses can actually use. This episode is for Mid-market SaaS founders tired of AI hype who want to build something that creates real customer value—not just impressive demos. Most SaaS companies don't fail because of bad tech. They fail because they chase hype over value. Zohar Bronfman, CEO of Pecan AI, took a different path. After years researching AI and philosophy in academia, he saw Amazon, Uber, and Spotify dominating with predictive AI—while thousands of smaller companies couldn't even get started. Instead of building another "AI for everything" platform, he focused obsessively on one thing: making predictive AI accessible to mid-market companies who couldn't afford data science teams. And this inspired me to invite Zohar to my podcast. We explore why curiosity beats strategy when building in uncertain markets. Zohar shares hard-won insights about deprecating profitable features, why small teams outperform large ones, and how to identify which enterprise capabilities actually matter for mid-market customers.