Providing quick product and design feedback via digital twins
TL;DR: Promi builds simulated versions of your real users from your product analytics data (Posthog, Stripe, Intercom, etc.). Run your next ideas past them: flows, copy, pricing, onboarding, and see where your users would hesitate, what they'd skip, and why, before anyone writes code.
Hi! Peter here, cofounder of Promi (S24)
The problem
Real user feedback is the gold standard for deciding what to build and almost nobody gets it, because it's slow and expensive. So teams substitute: survey panels, ICP interviews, and lately AI personas prompted with a demographic sketch. All of them share the same flaw — they don't know your users. The result is confident feedback with no context, and ideas that ship blind anyway.
How Promi works
Connect your product analytics tools. Promi builds behavioral profiles of your actual user segments, not personas, not survey panels, and turns them into simulated users. Describe a proposed change: a flow, copy, pricing, onboarding, and they react the way your real segments tend to: where they'd hesitate, what they'd skip, and why.
Why we're building this
I spent 5 years as a PM at Uber, where real user feedback was the gold standard nobody had time or budget for — so most ideas shipped blind. When we did get feedback, 9 times out of 10 we preferred existing customers over a generic ICP, because real users knew why they used the product. AI feedback needs that same context, so we built Promi to have it: simulated users trained on your customers' actual behavior.
Every product decision deserves real user feedback. We're making that the default instead of the luxury.
Our ask
If you're a product team with 3+ months of session data and ideas you're actively debating, reach out — we're onboarding design partners now. We'll build simulated versions of your users and run your next ideas past them. Then judge us on one thing: did it surface something true about your users that a generic AI couldn't? You know your users; you'll know fast whether our simulations ring true.
Schedule a call: https://calendly.com/pmoot/promi-overview
Website: https://www.trypromi.com Email: peter@usepromi.com
Promi - AI powered discounts for ecommerce merchants
How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?)
We both worked at Uber prior to Promi, where Peter led product for discounts across Uber Eats and Rides. Jiaxin worked on Uber’s core AI team. Peter had the idea for Promi after having seen how wasteful discounts were for the business, and how impactful certain features could be for making discounts more efficient and ROI positive.
