Data Decision Tool for Institutional Investors
Daily active users increased by 25 percent, exports to Excel dropped by 60 percent, and the redesign contributed to a 10 percent uplift in contract renewals.
-
B2B fintech analytics for institutional investors
-
UX focus on investigation instead of cosmetic dashboards
-
My role in UX strategy, product alignment, and leadership
About This Project​
​
This project focused on an institutional analytics platform used by asset managers and traders who live in data all day. The goal was not to make the dashboard prettier, but to design an investigation workspace that matches how they think about risk, exposure, and time-sensitive decisions.
​
The platform supports multiple institutional clients with billions in assets under management across global markets. I led the UX strategy end-to-end, partnering with Product, Data Science, and Engineering from discovery through rollout.
The Problem
​
Institutional investors were churning because the core analytics dashboard was slow to load and forced everyone into the same rigid layout, no matter their role or strategy. Analysts exported data to Excel to do the real investigation work outside the product.
​
We had strict constraints. Calculations needed to be accurate to four decimal places. Dashboards had to feel responsive with sub 200 millisecond load times. Data pipelines and permissions were tightly regulated. Users worked on multi-monitor setups in trading floors and research offices and expected the tool to respect their expertise and vocabulary.
​
-
Daily active users increased by 25 percent, exports to Excel dropped by 60 percent, and the redesign contributed to a 10 percent uplift in contract renewals.
Discovery and research
Designing for Trust in a Slow System
To understand why and when people used the product, I ran Jobs to be Done interviews with portfolio managers, traders, and risk analysts, and combined this with analysis of login patterns and export to Excel behaviour.
​
A clear pattern emerged. People did not log in to browse a dashboard. They arrived with specific, often time-critical questions such as whether sector exposure was safe ahead of a macro event, or which holdings would breach internal risk rules if an index moved a few percent. The tool needed to behave more like an investigation environment built around questions, not a static report.
Experience mapping and problem framing
I mapped the gap between how the system organized data and how investors organized their thinking. The platform grouped information by feeds, tables, and internal schemas. Investors thought in terms of strategies, risk buckets, exposure scenarios, and time horizon.
​
That mismatch slowed decision-making because users had to mentally translate raw tables into their own model each time. We reframed the product from a dashboard that shows everything into an investigation canvas that assembles the right views around a specific question. That reframing became the north star for all further design work.
How the platform stores data compared to how investors think about portfolios and risk. This gap drove the shift to an investigation canvas.
.png)
Information architecture and workflow
Based on the reframing, we designed a three level drill down.
Level one shows portfolio health and key risk indicators.
Level two shows sectors and themes so users can see where performance or risk deviates from expectations.
Level three shows individual instruments with history and scenario views so users can answer what-if questions in detail.
​
From any level, analysts can jump into filtered instruments or a saved workspace without rebuilding filters from scratch. For this expert audience, we kept professional terminology and reduced complexity through better grouping, smarter defaults, and more direct navigation paths instead of replacing domain language.
.png)
Three level drill down from portfolio health to sectors and individual instruments, with shortcuts into filtered lists and reusable workspaces.
Prototyping and validation
I created a reusable library of chart and widget components and assembled several workspace layouts in mid-fidelity. We ran desirability and usability sessions with risk analysts and portfolio managers, using realistic data and testing both light and dark contexts.
Early dark mode designs looked strong in Figma but failed on real trading floor monitors. Thin lines and small labels were hard to read, and some colour pairs did not meet accessibility standards. Before rollout, we rebuilt the color tokens and chart styles and validated contrast and legibility on actual client hardware in realistic lighting.
Light and dark tokens, typography, and chart components tuned for legibility on real trading floor monitors instead of just in design tools.
.png)
Impact and outcomes
The redesigned workspace changed how clients used the platform. Export to Excel actions dropped by about 60 percent, which showed that more of the investigation stayed inside the product. Daily active users increased by about 25 percent as analysts built and reused workspaces tailored to their own strategies. Net Promoter Score among institutional clients rose by 15 points.
On the commercial side, these behaviour shifts supported a 10 percent increase in contract renewals in that quarter. The analytics workspace became a central talking point in renewal conversations and helped strengthen the case for continued investment.
Export to Excel reduced, daily use increased, and sentiment improved, all pointing to the workspace delivering real decision-making value.
.png)
