I spoke on a panel at the AI Builders Forum this week about a question I’ve seen play out firsthand:

What does it really cost to do AI well?

At a packed meetup of AI founders and builders, my main point was simple: moving AI from prototype to production requires earning trust first.

I shared a lesson from one of my AI initiatives. Engineering wanted to move fast, but operations needed confidence that the AI recommendations wouldn’t create more work. Instead of rushing to scale, we prioritized precision and validation first.

That focus paid off. When users see accurate results early, adoption follows. Skip that foundation, and you may pay twice - once to build the technology and again to rebuild trust.

I also enjoyed the fireside chat, the other panel discussions, and meeting so many thoughtful people across the AI ecosystem.

Thank you to Khyati B. from eChai Ventures, PMI - San Francisco Bay Area Chapter, and AI Foundry for bringing us together!