Industry
Role
Scope
Problem
Triangle is a concept task-tracking app I designed through two UI challenges: a sign-up form and in-app search. Productivity tools fill up with tasks, documents and people fast. Both the first impression and finding things later often feel heavy. The goal was a calm, trustworthy interface that gets new users started quickly and lets them find anything in seconds.
Solution
A clean, blue-accented interface with clear hierarchy. The sign-up screen pairs a short form (Google sign-in plus three fields) with a product preview that shows real value: tasks, statuses and a productivity card. Search lives in a command-style modal with type filters (Tasks, Documents, Users, Actions), relevance sorting, results grouped by type and an Ask AI entry point for questions that aren’t keywords.
Result
Two core flows designed end to end, from empty state to results, on a shared set of components: status chips, task cards and document rows.
The search modal: filters by type, relevance sorting and results grouped into tasks and documents.
Insights
The first screen is a promise. People commit more easily when they can see what they’re signing up for, so the sign-up screen previews the product instead of a generic illustration.
In busy tools, search is navigation. Once an app is full of tasks and documents, people jump around by searching more than by browsing menus. Search had to be a first-class feature, not a box in the corner.
Not every query is a keyword. Some searches are really questions, like what’s due this week. That’s why Ask AI sits right inside the search bar instead of in a separate tool.
Process
I took two separate challenge briefs and turned them into one product, so the screens share a system instead of being one-offs.
I explored a different product first: a job-application tracker on the same layout. Then I refocused on task tracking, which gave search more to do.
I designed search state by state: empty modal, typing, then filtered and grouped results.
Search from start to finish: the empty state with the Ask AI entry point, then filtered, grouped results.
Key decisions
Results grouped by type. Tasks and documents get their own sections and layouts, so people can scan for the kind of thing they need.
Filters as tabs, not a dropdown. One tap narrows the results, and the active filter is always visible.
Status with icon and label. Chips like Completed or Scheduled never rely on colour alone, which keeps them readable for everyone.
An earlier direction: the same sign-up layout for a job-application tracker, before I focused the product on tasks.
Reflection
Treating two separate challenges as one product pushed me to think in components and states, not single screens. Next, I’d test search with real users and design the Ask AI answer state.
Available for new projects
Previous project