Inside Cohort 1, Week 1: helping famagic make family organisation feel simpler

Our first Beginner UX (AI) Design class introduced students to famagic, a start-up founded by working parents Yvonn and Henning Rau. Their app brings calendars, tasks, shopping lists and other family information together in one place, helping families organise everyday life and share responsibilities.

The founders used AI-supported development tools to build the first version, drawing on their professional backgrounds and their own experience of raising two children. This gave them a foundation they could put into the hands of families and continue developing with their feedback.

Early feedback highlighted opportunities to make parts of the experience clearer and more intuitive. During the beta phase, the team is using these insights to guide improvements. Areas for the students to investigate include onboarding, information hierarchy and how clearly the app communicates what happens after an action.

famagic is this cohort's client, and its founders are working directly with the students. Between now and their final presentations in December, students will use research, prototyping and usability testing to develop improvements to a selected part of the experience. The aim is to simplify everyday workflows, keep screens focused on what matters and use clear, supportive language that makes family organisation feel more manageable.

AI for UX, and UX for AI

Natalia drew a line early on between two things people tend to lump together.

AI for UX design is about how designers work: using AI to research faster, summarise, generate options and skip repetitive jobs.

UX design for AI is about what designers make. More and more products are AI-powered: a fitness app that builds your plan, a shopping app that recommends what to buy, a banking app that spots unusual spending and explains it. Designing those is different. The output can vary, so you have to design for the times it gets things wrong, show people what the AI can and can't do, and keep a human in control. Natalia's word for the hardest part was trust.

She also gave a practical way to choose an AI tool: what you need it for, your experience level, your budget, and, the one beginners forget, whether your client or employer allows it. You can't paste customer data into any tool you like.

Learning to see the detail

Then everyone screenshotted an app they love and dropped it into a shared FigJam board. We got a second-hand clothes marketplace, a travel planner, a language-learning app, a whiteboard app and a gym tracker.

This is where it got good. On the travel app, two screens from the same journey used different icon styles and colours. "It feels like one team worked with another team but never met each other," Natalia said. It's a small thing, but people read a different icon as a different function, so it adds confusion.

On the same app, a section title was set smaller than the content underneath it, so your eye jumped past it. On the learning app, she pointed to a progress bar that showed a percentage and the time left, a clear example of a system telling you where you are. On the gym tracker, she asked where your phone actually is when you're mid-set, on the floor or in a pocket, because context changes what a good screen looks like.

Her advice was to keep noticing. The more apps you look at closely, the more ideas you've got to draw on when you're designing.

Heuristic evaluation, and the AI rule

The theory half covered Jakob Nielsen's 10 usability heuristics. One that stuck was the "emergency exit": every action needs a clear way back. Natalia's example was the iPhone's shake-to-undo. Her other was GOV.UK, which asks one question per screen so you're never juggling five things at once.

Heuristic evaluation is cheap and fast. A usability test can take two or three weeks to set up; an expert review can start today. It doesn't replace talking to users, but it gives you a first list of problems to check.

The Week 1 project task is to evaluate the client's app against all 10 heuristics, then run the same evaluation through an AI assistant and decide which of its suggestions you agree with, and why.

The order matters, Natalia said. Think first, then ask the AI. If you ask first, all your thinking starts from its answer. She compared it to reading a book versus watching the film: when you read, you imagine it yourself. That's where your own ideas come from, and it's how you avoid designing the same product everyone else does.

Next week

Students are mapping the client's stakeholders, writing interview questions (business goals, users, problems, constraints and how success will be measured) and reviewing them with Natalia in their first mentor session. Then they'll interview the founders. Week 2 is user research.

Want to try this week's exercise on an app you use every day? Our free heuristic evaluation tool walks you through all 10 principles, and the stakeholder map is the one Natalia showed in class. For the bigger picture, read how AI is changing UX design, or see what's in the Beginner UX (AI) Design course.