UI/UX +
Visual designer

The challenge

I needed to test how different AI tools work together in different situations. Could Claude Code read Figma files accurately? Could it also create a design system in Figma? How would Claude Design fit into the process? And could I vibe-code a feasible MVP using version control, so the app would be ready for handoff to development?

My role

Concept & prompting
UI + Visual design
Prototyping
Front-end build
Version control
Accessibility audit

Results

• Claude Code could read a Figma design system directly and build a working prototype from it, then push the coded version back into Figma as a new design system
• Claude Design coded fresh visual directions in minutes, but could not share the results directly with other platforms
• I turned that exploration into a real, accessibility-audited product with Claude Code
• I could hand the finished prototype off through GitHub the way a developer team would: versioned, branched, pull-requested, and live on a public, auto-deploying address

From a Figma design system into Claude Code, out to Claude Design for new ideas, and back into a working MVP handed off through GitHub: a versatile AI-built chain that I can harness for different work situations. Of course, the scene changes so rapidly that a couple of months from now I’ll most likely have more tools to test and use.

Full case study: Reshaping Workflows

The challenge

Create a digital service that brings housing association data into one clear, easy-to-use app. The goal: improve transparency for residents and save time for property managers.

My role

Research
User flows
Wireframes
UI + Visual design
Style guide
Accessibility
Prototype

Results

• Apartment owners access financial data, documents, and messaging in one place
• Accessibility focus: large touch targets, high contrast, simple navigation
• My research revealed older than expected user base
• No usability issues reported post-launch
• Oiva gained a competitive edge with transparent, data-driven property management

A calm, accessible app that builds trust, supports aging users, and sets Oiva apart in their industry.

Full case study: Transparent living

The challenge

Design a tool that shows ferry captains how to drive more efficiently: when to use power, when to hold back, and how close they are to the AI-calculated optimum.

My role

Research
User flows
Wireframes
UI + Visual design
Accessibility
Prototype

Results

• Emissions and energy use decreased 28%
• Captains learned efficient techniques in real time
• Simple, game-like interface: match your curve to the “perfect” one
• Leaderboards tested, but dropped. Too few captains, morale impact

A serious B2B tool that feels like a game. It’s teaching efficiency, saving money, and cutting emissions.

Full case study: Gamifying Efficiency

The challenge

How do you grow a simple tool for tracking hours and photos into a full ERP platform that covers mapping, documents, budgets, salaries, ticketing, deliveries, and warehouse management, while working with a very small team and limited resources? The system also needed to support the company as it scaled rapidly, ensuring that new processes and a larger workforce could still run smoothly.

My role

User interviews
User flows & wireframes
UI & visual design
Style guide
Accessibility
Prototype
Feedback survey

Results

• The ERP grew into a versatile platform that adapts to different user types and devices
• The system enabled the company to scale its operations significantly while keeping processes efficient and manageable
• Survey feedback showed high usability, with some room for more polish in quality and coherence
• A long backlog of feature requests reflects both strong adoption and ongoing business growth

Built by a tiny team, this ERP became the backbone that allowed the company to grow far beyond its original size.

Full case study: Evergrowing ERP

The challenge

Design two AI-powered bots: one to help staff process customer forms and one to guide the public on the website. Both needed to deliver quick, reliable answers, but with technical limits on freeform conversation the UI had to set clear expectations and build trust.

My role

Research
User interviews
User flows & wireframes
UI & visual design
Style guide
Accessibility
Prototype

Results

• Two tailored bots: search-like interface for staff and chat-style for the public
• Early feedback from users and client is positive
• Design insights gained on expectation-setting, trust, and the role of copywriting in chatbot UX

By setting realistic expectations and focusing on trust, the AI chatbots open a path toward better digital support for both citizens and employees.

Full case study: Setting Expectations

Design brief

Create a digital service that streamlines the leasing experience:

• Easy access to contracts, schedules, and bike specs
• Renewal reminders, support, and forms for different scenarios
• Insights for the leasing company and partner shops
• Tools for employers to track benefit usage among staff

The app should feel approachable, sympathetic, and easy to use.

My role

Research
Concept design

The concept

• Visual style based on the company’s brand: large text, vector images, vibrant colors
• Adjustments for accessibility and readability
• Dark blue backgrounds for depth, with white boxes for clarity
• Rounded shapes to add friendliness and approachability
• Potential use of subtle animations to enhance engagement

Next steps

This concept establishes a foundation for a welcoming, accessible leasing platform. With refinement and testing, it could evolve into a service that strengthens the company’s brand while making bike leasing easier for everyone.