Designing complex, data-heavy systems for enterprise software companies

How I approach enterprise software design

Design for batch workflows, not single tasks.

Enterprise users work in stacks. Technicians enter claims in a single session. Analysts process documents in batches. Security technicians triage multiple issues at once. The flow should be optimized for rapid sequential entry, not one-off submissions. At Subaru, that meant starting a new claim in two clicks. At Optum, that meant a history table that let analysts jump back into any prior process.

Show bad news first.

Problems should surface immediately, not at the end of the workflow. Inline validation prevents wasted work and builds trust. At Subaru, moving validation from the end of the claim to the point of entry reduced rework and sped up processing. At Optum, flagging failing documents at the top of the queue let analysts triage the most urgent issues first.

Build the system, not just the screens.

A single dashboard is not a design solution. It is a component of a larger system. Design tokens, reusable components, and consistent navigation patterns make the next feature faster to build and more consistent. At Optum, extending the Harmony design system with AI-specific components made every subsequent AI feature faster to ship. At Subaru, building a standardized Material Design system was necessary to ship the MVP on a short timeline.

Pressure-test your own designs.

Enterprise software fails in edge cases. Unusual values. Missing data. Unexpected user behavior. I used Microsoft Copilot to audit dashboard designs for edge cases I might have missed. The best way to catch blind spots is to actively look for them.

Validate with the people doing the work.

Stakeholder proxies are useful for alignment. They cannot replace watching real users work. At Subaru, I interviewed approximately a dozen current and former technicians. At SSA, I interviewed field office managers directly. At Optum, I partnered with the Senior Director of Architecture and the Senior Technical Project Manager to define what success looked like before designing anything.

This is the same problem I solve across federal, analytics, and enterprise software: complex, data-heavy workflows where a slow interface directly impacts operational cost and customer experience.