Continuous Discovery at Scale: How AI Can Transform User Research
Tomorrow, May 21st, at 7 pm BRT, I’m running a free masterclass on one of the most important — and most neglected — topics in product management: how to talk to your users continuously, deeply, and at scale.
This talk was originally prepared for PM3 Summit, one of Brazil’s largest product conferences. My flight got canceled, and I couldn’t make it. So instead of letting the content sit, I’m opening it up as a free online session.
Why this matters
AI has already transformed how product teams build software. Vibe coding, AI-assisted design, faster prototyping — delivery has never been more accessible.
But there’s a bottleneck that almost no one has solved: talking to users. Continuously, deeply, and at scale.
Recruiting, scheduling, interviewing, transcribing, analyzing. The process is too heavy to become a habit. So most teams do discovery when they can, not when they need to. And as AI raises the bar on delivery speed, knowing your users becomes even more critical — because the risk of building the wrong thing faster has never been higher.
What I’ll cover
How AI agents are changing the dynamics of qualitative research, making user interviews more frequent, more accessible, and integrated into the daily routine of product teams — without relying on external recruitment, complex tools, or one-off research sprints.
I’ll show, in practice, how WhatsApp can become a live channel for continuous discovery and how to turn user conversations into actionable insights at scale, with less friction and faster learning cycles.
At the end, I’ll demonstrate how this approach is being applied today, drawing from real experiments and a product built specifically to solve this problem.
Tomorrow, May 21st · 7pm BRT · in Portugues
Free Register here: https://forms.gle/et8DnJufwtQPLs9k7
Not a Portuguese speaker?
If you’re interested in an English edition of this masterclass, register your interest here: https://forms.gle/T3wHR85VjoC61q7WA
I’ll reach out if and when it happens.