Conversational AI Design: A Practical Field Guide
Details
About the talk
Conversational assistants aren’t “just another feature”. They are full experiences shaped by countless design decisions that must hold up in real life: messy inputs, high expectations, edge cases, and the constant tension between speed, quality, and cost.
In this talk, Gea and Alberto share what they’ve learned while designing AI-enabled assistants in different contexts (from support to retail-like journeys). Instead of focusing on hype or theory, they’ll unpack the practical choices that make the difference: how to keep conversations actionable, how to design a consistent “voice” without overdoing personality, how to define (or intentionally limit) the assistant’s identity, and how the interface and the underlying logic need to evolve together.
They’ll also touch the less glamorous, but most decisive parts of the work: when AI is the right tool (and when it isn’t), why data is the elephant in the room, what “testing” means in conversational design, and how to guide users so the assistant is easy to understand and hard to misuse without overwhelming them.
Expect a set of clear takeaways you can reuse with your own team, whether you’re starting from scratch or improving something already in production, shared through real design trade-offs, not spoilers.
About the speakers
Gea Sasso is a Senior Service Designer at Sketchin focused on service experiences where journeys, touchpoints, and contexts come together. Her work explores how services can be shaped across channels, including the interplay between digital and physical environments.
Alberto Andreetto is a Lead Interaction Designer at Sketchin specialized in interaction design and complex digital systems. He works on defining how products behave, respond, and communicate, especially when interfaces become conversational and adaptive.
Sketchin is an advanced strategic-design studio that helps forward‑thinking organizations explore technology and face their next sustainable change. They combine business opportunities with customer expectations to design and deliver end‑to‑end experiences, often leveraging deep tech and AI to reshape products, services, and customer engagement.
---------------------------------------------------------------------------------
💬 The Talk is in English
📍 *Note* The event is happening in a new room *note* The event will be held at ETH Zürich Maschinenbau ML, Room F34
We kindly ask you to arrive 10 minutes early of the event start, so everyone can find a seat due to high attendance. Thank you!
