Fintech & trading
Finance interfaces where a model’s output has to be trusted in seconds.
Dense, multi-role workflows on desks and in apps, where speed, interpretability and compliance pull in different directions.
What’s different here
- 01
Density is the job. Portfolio managers, analysts, traders and sales each need a different cut of the same data, now.
- 02
Interpretable or ignored. A machine-learning signal a trader can’t read, question and act on is noise.
- 03
Latency is UX. Monitoring, alerts and analysis have to feel instantaneous.
- 04
Regulation shapes the interface. Audit trails, permissions and explainability are on-screen requirements.
- 05
Trust is earned per transaction. Onboarding, verification and money movement live or die on clarity.
What we’ve done
Led UX for a fintech’s machine-learning relative-value bond-trading platform: contextual research with portfolio managers, analysts, execution traders and sales, then features across search, analysis, monitoring and visualisation.
Designed the AI-powered features for detecting relative-value dislocations, and the interaction patterns that made model output interpretable and actionable for domain experts.
Built, on our own product, a payments and credit system with microcent-level cost tracking, and a multi-provider AI architecture with rate limiting and failure recovery.
What we bring
- Data-heavy UX and information architecture
- AI/ML interface design (confidence, provenance, override)
- Research with specialist users
- Design systems for dense interfaces
- Payments and onboarding flows
- Compliance-aware product definition
Principles
Show the signal, the confidence and the source, or don’t show the signal.
Every alert costs attention; spend it deliberately.
Speed is a feature; clarity is the feature.