Building AI-Native Products: A Design + Engineering Playbook

Isa Simó and Miquel Farré first learned to collaborate at Disney: product on one side, engineering on the other. Now they’re reunited at Studio Jadu as Head of Product Design and Head of AI, building products where AI isn’t an add-on but the foundation.

A lot changed! LLMs are making thiner and thiner the boundary between UX and AI. When the interface is a conversational assistant (capturing intent, asking questions, generating artifacts, taking actions) the conversation is the product. This breaks fundamental assumptions: deterministic flows, static screens, predictable edge cases. In this talk, we share a joint design + engineering playbook for building AI-native products that remain usable and trustworthy even when probabilistic.

We’ll explore designing when the primary UI is communication itself: structuring intent, maintaining orientation, making outcomes legible. Then we’ll connect this to production engineering: context management, guardrails, evaluation, observability, and safe action patterns.
Finally, we’ll share how we iterate rapidly without shipping chaos—through tight loops with real workflows, measurable quality signals, and processes that align design with technical constraints.

You’ll leave with: patterns for assistant-first UX, a practical design + ML iteration framework, and concrete tactics for keeping AI-core products reliable as they evolve.