Portfolio management

Challenge

Our previous portfolio management feature was built to meet table stakes: it worked like a transaction ledger, simply put. Customers were coming to us with much more complex planning needs. They needed to understand inventory across delivered, undelivered, and not-yet-issued credits, then decide which credits to retire against specific future years and claims. The complexity came from the lifecycle itself: issuance, delivery, payment, and retirement all moved on different timelines, and the product wasn't helping customers make sense of that.

Customer Research & Framing

We treated this as a focused generative research effort over about three weeks. I put together the research plan, got buyer interviews on the schedule, and combined internal conversations with Patch teams and direct customer interviews to understand how people actually thought about inventory, status, and retirement planning. The first week was planning, the second was execution, and the third was analysis of the insights, all that while pushing the actual design explorations forward in parallel.

Designing for Complex Inventory

The core design challenge was making a complicated portfolio feel legible without flattening the nuance or losing the details. Customers needed to see what had been purchased, what was actually available, what was still pre-retirement, and what had already been allocated or retired. I focused on organizing the experience around planning and traceability, so buyers could understand both the current state of their inventory and what actions they could take next. That made the system more useful for day-to-day decision-making and for reporting to investors or governing bodies.

Build & Collaboration

This project moved in a very fast build environment, with engineering using AI heavily to accelerate execution. My role was to create enough clarity that engineers could keep moving without waiting on every final UI detail, while still making sure we weren't under-designing and shipping something too loose. We worked in a tight loop where prototypes, specs, and implementation evolved together, and I made sure the team had a strong enough view of the direction that they could stay unblocked as the interface firmed up. That balance between speed and product judgment was one of the most important parts of the work.

Outcomes

This wasn't a paid feature, so success showed up in operational impact rather than direct revenue. After launch, we saw an 82% decrease over three months in customer requests to sales & operations asking basic but critical questions about their credits and status.