← All work FINTECH · DATA VISUALIZATION · INTERNSHIP

Fidelity Investments

Redesigning risk analysis tools for portfolio managers - people whose job is reading complex financial data quickly and acting on it. The work centred on one question: how should two separate tools connect to each other?

Role: UX Designer, Intern
Team: ATIM - 6 people
Timeline: May 2024 - Aug 2024
Scope: Research, IA, prototyping, usability testing

~13%Task efficiency gain

5 → 2Flow options tested down

2Tools redesigned

The problem

Portfolio managers at Fidelity used two tools to understand risk. PMRI showed where a portfolio stood against its benchmarks. Scenario MRI let them model what would happen under different market conditions.

The tools worked. What didn't work was the relationship between them. A manager looking at a risk indicator and wanting to model it had no clear path from one to the other - the analytical thought was continuous, but the software made it a stop and restart.

Alongside the redesign, the team wanted a concept for a new web app aimed at premium brokerage users. Different audience, different needs, same three months.

Key decisions

Decision 01

Testing five ways to connect PMRI to Scenario Builder

The tension: There was no obviously correct way to link the two tools. Tighter integration meant a smoother path but a heavier, more cluttered PMRI. Looser integration kept each tool clean but left the manager to bridge the gap themselves.

Rather than argue the options in the abstract, I built all five and put them in front of the people who would use them. Each represented a different position on that spectrum - from launching Scenario Builder as a separate destination, through to embedding scenario editing directly in the PMRI view.

Think-aloud sessions with portfolio managers narrowed five to two. What separated them wasn't preference; it was where people expected to end up after clicking. The options that survived matched the managers' existing mental model of the analysis - the ones that didn't required them to learn a new one, and that showed up immediately as hesitation.

Decision 02

Accessibility as a constraint on the charts themselves

The tension: Financial visualisations lean heavily on colour - red for loss, green for gain, colour-coded risk bands. That convention is deeply held in finance, and it excludes anyone who can't distinguish those colours.

I treated non-colour encoding as a requirement from the first wireframe rather than a compliance pass at the end. Status carried in labels and position as well as hue, contrast ratios checked as the designs developed, and interactions built to work from the keyboard.

Doing this early cost some visual density in the first drafts. It also meant the charts didn't need reworking before engineering handoff, which on a twelve-week internship was the difference between shipping the redesign and handing over an unfinished one.

Outcome

Both tools were redesigned and handed to engineering: PMRI with clearer graph hierarchy, better tooltips and a more legible call to action, and Scenario MRI with the connection model that came out of testing. The premium brokerage web app was delivered as an end-to-end concept with custom widgets.

Usability testing measured roughly a 13% improvement in task efficiency against the previous tools. The more useful signal was qualitative: managers stopped asking where a control had gone, which was the most common complaint at the start.

What I'd do differently

Five options was probably two too many. Three of the five were eliminated quickly and for reasons I could have anticipated with a sharper read of how managers already worked. Building them cost time I could have spent testing the surviving options more deeply.

The instinct - test rather than argue - was right. The calibration was off.