Quantifire attention overview dashboard
Overview
Product
Data Visualisation
AI Powered
Design System
Conclusion

Quantifire

Quantifire
Product · B2B Intelligence/2026 — ongoing/In build

Overview

Quantifire’s investor intelligence arrived as static PDF reports. I used them as a source of information, then designed a desktop product that let people explore that information rather than follow a fixed narrative.

Scope: Investor dashboard and data visualisation, account and relationship views, data states, the design system across light and dark, and Iris, the in-product AI layer.

The reports were thorough but fixed: revenue, exposure, holdings movement and sentiment were all in there, but the structure dictated what people saw first. My work was to decide what each audience actually needs, prioritise and structure that information, and design interactions that let them explore rather than follow a fixed narrative.

So the product does two things static reporting could not: move users from overview to detail in the same place, and let them interrogate a figure instead of taking it on trust.

Impact

  • 01Rebuilt static investor reporting as one interactive desktop product
  • 02Turned conventional charts into a considered visual language for revenue, exposure and portfolio movement
  • 03Made currency, variance and time comparison consistent across every view
  • 04Designed empty, partial, stale and low-confidence states as the normal case
  • 05Reserved colour for meaning: gain, loss, exposure and risk
  • 06Made Iris, the in-product AI, checkable by citing the records behind every answer

The Product

Account overview, primary desktop view

One surface, three audiences

Depth stays available, but the opening view answers the question each role asks first. Layers are configurable rather than duplicated, so one dashboard serves analysts, account leads and sponsors.
Relationship and sentiment view

Signal over volume

Investor relationships, momentum and risk are surfaced visually before anyone opens a table. Quieter detail sits one interaction away, so the overview stays a route into the data rather than a wall of it.

Complexity wasn’t the problem. Undifferentiated complexity was.

Dashboard, full desktop screen
Detail view, drilled-in revenue data
Table view, dense financial data states
Attention analytics, visitor and engagement overview

Hierarchy

The dashboard opens on the holdings that moved, not on every holding at once.

Density

Investor tables carry more rows because type, spacing and colour carry less.

States

Partial, stale and low-confidence figures are designed, not handled later.

Data Visualisation

The goal was not prettier charts, but clearer signals from complex information.

Revenue and exposure over time

Money movement, first

Standard charts were rethought around the investor question: what changed, by how much, and against which period. Variance and direction read without a legend, so the visual treatment serves the signal.
Portfolio breakdown and variance

Comparable by default

One currency treatment, one number format, one time-comparison pattern. Figures stay comparable across portfolio, account and line-item views, so relationships and changes are easier to recognise without re-reading each axis.

Gain, loss, exposure

Colour is reserved for financial meaning, so gain, loss and exposure read instantly and consistently across every view.

Precision on demand

Rounded figures keep the overview scannable; exact values appear on hover or in the export, so detail is available without overwhelming the first view.

Confidence shown

Modelled, lagging or low-confidence numbers carry their own visual treatment, so they are never mistaken for settled data.

AI Powered by Iris

Ask the numbers a question, not a query.

“Why is Q3 revenue down on this account?” is an investor question, not a filter. Iris, the in-product AI, sits over the same data the charts use, so an answer arrives as figures, movement and drivers, with the records that produced it attached.

The design problem was trust rather than intelligence: an AI answer about money is only useful if the reader can open the numbers behind it in one click, and see when the model is inferring rather than reporting.

Iris, in-product AI panel over the financial view
Iris answer with cited financial records
Project home with Iris chat entry point
Iris suggestions, analysed and prioritised
Project knowledge feeding Iris answers

Prompt

Natural language, scoped to the account and period already on screen.

Answer

Structured, not prose: the figure, the variance, the driver behind it.

Provenance

Every number links back to the holdings and records behind it.

Honest limits

Modelled or lagging answers say so, and Iris declines rather than guesses.

The Design System

Type & density scale

Type & density scale

A restrained ramp with two data-specific sizes, so dense holdings tables and exec summaries share one voice.
Data components

Data components

Cards, tables and charts specified once, with empty, partial, loading and low-confidence states built in.
Colour as signal

Colour as signal

Colour tokens carry meaning, not decoration, so nothing competes with sentiment, movement or risk.
Light and dark, dark mode
Dark mode
Light and dark, light mode
Light mode

Light and dark

The same components and tokens resolve into both modes, holding hierarchy, contrast and readability steady wherever an analyst works.

Conclusion

Fixed reporting became a product users can interrogate, and a system dense enough to grow into.

Hierarchy became a product decision rather than a styling one. Early versions mirrored the original document page by page; later rounds cut charts, added states, and let each role land on the view that answered their first question.

Key decision: kept Quantifire desktop only. It cost reach, but it bought the room investor data at this density actually needs.

Other case studies

Tribe

Product · Community · 2025—ongoing

Memberships, events and ticketing as one connected product.

View

Maybourne

Website · Hospitality · 2025—26

IA, UX/UI and a design system across six luxury properties.

View

OVO Energy

Mobile App · Everyday · 2024

An energy app built around the questions people actually ask.

View