Memi Beltrame

Artificial intelligence is more and more becoming the core of digital products. Designing for Products based on AI requires Designers to know about Machine Learning. This talk is an easy walk through the most important elements of Machine Learning. It looks at the fundamental principles of using practical examples. It showcases applications of the different types of Machine Learning. The use-cases range from text categorization to image recognition, on to speech analysis. The goal is to show what is important for designers and why.

Language is a vital part of design, critical to the design itself but also to the process underneath. It affects companies on many levels: different roles, backgrounds and languages spoken within the team bring the need of a common vocabulary. Companies set up a vocabulary of some sort but often struggle to make it a part of the company culture. This talk looks at the challenges of controlled vocabularies and how they can be helpful in creating a shared understanding of the company.

An in-depth look at how interaction design, industrial design and service design are merging together to form a super-discipline and what this all means for designers.

This is a case study of 3 projects with a seemingly trivial task in common: how do we show users meaningful suggestions of what they can buy in a web store without knowing much about them? The first project is about us struggling to design meaningful suggestions using conventional patterns. In the second, we shifted to a conversational UI and saw its mechanics change the dynamics of interaction. In the third project, we had an epiphany where it became apparent just how much designing conversational UIs is tied to understanding language, speech acts and the psychology of human behaviour.

Modern web design often focuses too much on delivery and not enough on function and data. This talk advocates for data-driven design (DDD) that simulates user interactions with actual data. The presentation covers the importance of realistic prototypes, introduces Protostrap as a prototyping tool, and demonstrates patterns for handling large amounts of data including filtering, bulk editing, and assignment workflows. The key message: Users love their data, but don't relate to dummy data.

Traditional A/B testing divides users into static groups, which can be costly when large user numbers are needed. Dynamic A/B testing addresses this by giving 10% of users a random option and routing 90% to the best performing option based on real-time metrics (views, clicks, click-through rate). This approach allows for faster conclusions with fewer users and significantly reduces opportunity costs. The talk introduces AB/CD, a service that handles the mathematical calculations behind dynamic A/B testing through a simple API.

This talk explores how Near Field Communication (NFC) technology can transform everyday interactions by connecting the physical and digital worlds. It addresses three main pain points: cumbersome interfaces, unnecessary manual tasks, and lack of context awareness. The presentation demonstrates practical applications in travel, healthcare, and home automation, showing how NFC enables seamless experiences through smartphones that act as personal readers. Key message: NFC represents a shift toward an Internet of Things where objects have identity and can communicate, making manual processes digital and context-aware.

This lightning talk uses Zurich's ingenious dual water supply system as a metaphor for designing resilient systems. Zurich maintains two completely separate water networks - one for daily use and decorative fountains that serve as an emergency backup. These beautiful fountains can be opened to reveal hydrants for water distribution in disaster scenarios. The key takeaway: Always design for failure, worst-case scenarios, and edge cases - whether it's no JavaScript, low bandwidth, flaky connections, offline mode, or inconsistent user input. Plan for disaster from the start.