Runroom LAB Designing for Humans and Agents

The journey towards an AI-native design system. The Manychat case

Format:: Online
Schedule:: Thursday, November 12, from 6 PM to 9 PM
Location:: Runroom (in-person + streaming)

Everyone talks about AI in design. Few share what really happens when you use it seriously.

At Manychat, they were pushed towards AI. They went all-in as an "AI-active" design team and then, surprise, ended up back in Figma. This talk is the honest story of that round trip: what they tried to solve, what broke along the way, and where AI earned its place.

Told from both sides, design and engineering, Agostina and Thành will show how they rebuilt their design system to be AI-native: a cross-platform base (Web, iOS, and Android) where designers, engineers, and AI agents work from the same source of truth.

Who this Runroom LAB is for

Product designers, design system teams, web and mobile engineers, design leads, and product managers who are already working with AI (or are about to) and want to go beyond the hype.

Key takeaways

How an AI-native design system changed their way of doing discovery and delivery.
A concrete case of how the system accelerated adoption and delivery speed.
A short demo of how they prototype with the system.
Honest lessons on when to rely on AI and when to return to the canvas.

Share:Linkedin/Bluesky

Speakers

Agostina Lotierzo

AI Design System Designer at Manychat

Originally from Argentina, I’ve been living in Barcelona for the past five years. Although I graduated with a degree in Graphic Design from the University of Buenos Aires, my career path led me to become a Product Designer, working with companies like Santander, Despegar, and eDreams. That journey brought me to my current role: AI Design System Designer at Manychat. My work is hard to summarize in a single sentence: I design the system that other designers, engineers, and AI agents use to build product—all from the same source of truth, cross-platform (Web, iOS, Android). I live between the canvas and the code: writing specs, reviewing PRs, and advocating for AI when it genuinely solves problems while remaining skeptical when it’s just hype. This is the fifth design system I’ve contributed to, but the very first one where machines are also users! I'm also an illustrator—a passion I rediscovered as I got closer to AI. The more automated things get, the more I return to creating by hand.

Thành Đỗ Long

iOS Developer en Manychat

Ingeniero iOS en Barcelona, con formación en tecnologías de la información y diseño gráfico. Empezó en el diseño y pronto dio el salto al desarrollo iOS: de la agencia Cleevio a empresas de producto como Ocado Technology y N26, hasta llegar a Manychat, donde iOS es su dominio principal dentro de un design system compartido entre iOS, Android y web. Trabaja entre el producto, el diseño y la ingeniería. Escribe especificaciones, revisa pull requests y construye la capa agéntica del sistema para que el código asistido por IA llegue a producción de forma segura, con decision records, workflows repetibles y CI gates. Le interesa la IA que resuelve problemas reales: menos trabajo repetitivo, más productividad y mejores productos. Nada de generar output por generar ni de lanzar una granada de slop a quien le toque limpiar: lo que busca es crear mejores sistemas y entregar mejor trabajo. Ganó el Swift Student Challenge de la WWDC21 de Apple y asistió a la WWDC26. Lo que más le sigue motivando son los detalles del día a día: que una interacción se sienta natural y que un código sea más fácil de evolucionar.

Reserve your spot

This Runroom LAB is free. You just need to register to secure your spot.