Career Foundry UI Specialisation | Student Project
An AI assisted property search experience that helps first-time investors make confident, informed decisions.
First time property investors want to make smart decisions but property searches are often overwhelming and leave people with unanswered questions . Perfect Properties reimagines the process through a user-first lens with clean design, AI support and an interactive map for a visual search experience.
Overview
Design Brief
Design a responsive, AI-assisted web app that simplifies property discovery for first-time investors seeking financial stability or additional income.
Partial Brand Guidelines
Scope
Student Project.
Timeline
Discover
The prompt I used to conduct user research in ChatGPT
I had brand guidelines, one persona and was asked to start straight into the UI design. However I needed more information to truly understand what first time property investors needed from this design. To gain insight quickly, I conducted research on ChatGPT and used the CARE (Context, Ask, Rules, Example) framework to enginer my prompt. I looked up the top 2 competitors, performed heuristic evalutions on their UI to ensure I didnt make the same mistakes and looked for a gap in the market by analysising their user reviews.
Explore
Balancing function and form on mobile UI
I sketched out user flows in detail, I realised that property searches were complex, they demanded a broad scope of information from the user and squeezing that into mobile design was trickier than I had anticipated.
Homepage v1
I tried to bring search results and educational resources together to support first time investors via horizontal scrolling, but this made the design extremely cluttered. So I decided to create 2 separate screens instead, a dedicated search results screen and a separate homepage that would act as a hub for personalisation and education.
Search Results
I catered to 3 core user needs on the search results screen by adding a FAB for AI chat support, a tool bar offering filter, map location, notification options, save features, and ‘key word’ quick filters which was earlier identified as a gap in the market,
Homepage v2
On the homepage I included a ‘updates’ feature to prioritise users 'notification' choices and entry points to additional features that would lead to further engagement, via a visual search map, personalisation and education.
Tabbed filters brought clarity and scalability
I looked at mega menu, accordion and a tabbed layout to house filters, in order to decide which one would reduce visual clutter, provide clarity and afford clean scalability. I felt tabs aligned with Perfect Properties brand guidelines of being clean, quick, and smart. Preference testing further validated that the tabbed design felt most intuitive and least visually overwhelming to users.
Desktop Homepage v1
The persona Rashida wanted to see property locations at a glance. I designed an interactive map to help her visualise her property search. Development for this feature would possibly be time consuming and expensive but making it accessible via the homepage would make it a strong value proposition for the business as it sets them apart from competitors.
Desktop Homepage v2
I started adding each feature to the homepage and soon, the layout became a war zone between branding and usability. A card sorting workshop would have helped me prioritise the structure but unfortunately time constraints meant this wasn't possible. Desktop traffic is strong for property buyers and careful consideration was required to strike the correct balance.
Desktop Homepage v3
I improved branding by selecting more complimentary imagery. Hierarchy was improved by using a horizontal layout for all elements. I invited users to learn by expanding ‘Top Tips’ feature to show the buying process and kept the AI feature accessible by using a FAB.
Build
Style Guide
I created the visual language for Perfect Properties from scratch to develop my skills as a UI designer. Making my own icons and assests was fun! Althugh using premade UI kits is faster for real world projects, I'm happy to be able to create custom pieces where necessary.
Deliver
Retro
What I learned about designing for people, not pixels.
Design AI around trust
Users connect with AI when it feels safe, clear, and helpful. They need to understand what AI can and can’t do. Every design decision, from microcopy to visual tone, shapes trust. This project strengthened my confidence in designing ethical, accessible, and human-centred AI interactions.
Adapting research
Working with minimal data taught me to use AI tools like ChatGPT and Perplexity for structured, rapid insight generation. This turned ambiguity into actionable design direction.
Clarity over quantity
Iterating through multiple filter layouts reinforced that good UX isn’t about showing more options, it’s about reducing cognitive load through clean, structured hierarchy.
