Navigating massive data without loading it all
Institutional traders need to explore enormous volumes of position data, often without knowing exactly what they're looking for until they start investigating.
Loading everything isn't practical. Neither is forcing users to formulate a precise query before they can explore.
The challenge was to design an experience that helps users move from a vast universe of data to the positions that matter, balancing information density, flexibility, and performance.
Role: Senior UX/UI Designer
Scope: Information architecture · Data-heavy UX · Interaction design · Scalable data exploration
Collaboration: Product Manager · Engineering · Domain experts
Scope: Information architecture · Data-heavy UX · Interaction design · Scalable data exploration
Collaboration: Product Manager · Engineering · Domain experts
Starting with the bigger picture
Users shouldn't need to know what they're looking for before they can find it.
A user might start by reviewing exposure across multiple books, notice an unusual concentration in a particular sector, and investigate the positions contributing to it.
But exploration at this scale presents two competing challenges: users need the flexibility to navigate enormous datasets, while the application needs to avoid retrieving information they may never look at.
The experience needed to support three complementary activities:
Define the universe: Select the accounts, books or positions relevant to the investigation.
Find what matters: Organize and aggregate the data to identify meaningful exposures.
Investigate further: Drill into individual positions without losing the broader context.
The challenge wasn't simply making a large table easier to read. It was making an enormous dataset explorable.
Aggregate first. Investigate second.
Why load thousands of positions when a handful of summaries might reveal where to look? Rather than presenting every security immediately, the experience starts with aggregated data.
Users can organize positions by different dimensions—book, sector, or asset class—and progressively expand the hierarchy to investigate what's underneath.
This approach makes aggregation more than a reporting feature. It becomes the primary navigation mechanism. Users can start with the big picture, follow what catches their attention, and progressively retrieve more detailed information.
Making the data model understandable
Two levels of filtering. Two different effects on the data.
On-demand retrieval introduced an important UX challenge. Query-level filters determined which data was retrieved and aggregated, while position-level filters applied only to positions already loaded in the grid. Filtering individual positions did not change the aggregated summaries.
For users unfamiliar with the underlying model, that distinction wasn't necessarily obvious. A filtered grid could display fewer positions while the summary totals remained unchanged, potentially creating confusion about what the numbers represented.
Much of the design work focused on making the scope and consequences of each control clear through their placement, hierarchy, and interaction patterns, without requiring users to understand the technical architecture.
The challenge wasn't simply distinguishing two types of filters. It was preventing users from misinterpreting their results.
A position-level filter could reduce the number of visible rows without changing the aggregated totals. The controls, their placement and the feedback around them needed to communicate that distinction naturally, without requiring users to understand the underlying data-retrieval model.
Designing around performance
At this scale, performance isn't simply an engineering concern. It directly influences how users navigate and interact with information.
The information architecture needed to prioritize meaningful summaries, support progressive exploration, and avoid retrieving unnecessary detail.
This meant designing around what users needed to see first, what could remain aggregated, and when individual positions became relevant.
The objective was to make useful information available early without loading everything upfront, while preserving the flexibility to investigate further.
Making density useful
Less information isn't always better. Better-organized information is.
Institutional users need to compare multiple financial metrics, scan large tables, and investigate anomalies without constantly switching views. The interface embraces density while establishing a clear hierarchy between aggregated exposure and position-level detail.
Configurable grouping, expandable hierarchies, and contextual information allow users to move between different levels of data within the same workspace.
Summary rows help identify where to investigate. Expanded positions provide the detail needed to understand what's happening.
The goal isn't to simplify the underlying financial complexity. It's to make that complexity navigable.
From data overload to focused investigation
The central design decision was to make exploration progressive rather than expecting users to search or navigate an entire dataset at once.
Define → Aggregate → Identify → Investigate
By bringing aggregation, filtering, and progressive disclosure together, the experience supports a continuous path from broad exposure to individual positions, while accounting for the technical constraints of working with large datasets.
The objective is simple: help users find what matters without looking at everything.