A five-year direction, not another set of payment screens
The question was how AI could improve the way people move money through a bank-owned platform. We needed a direction for the next five years, not another set of separate payment screens. My job was to help turn that opportunity into an experience people could understand and a proposal leadership could evaluate.
One role, two tracks: design quality and product direction
I led the consumer phase as the sole senior designer, working with a researcher and one other designer. The role ran on two tracks. One was design: holding the quality bar across a multi-designer sprint and keeping the concepts consistent enough to test as a single set. The other was product direction. I framed the questions, sequenced the phases, and kept stakeholders aligned.
Research decided the MVP
I started with the competition. Across six core money-movement use cases, I tore down more than fifty products to find the patterns worth keeping and the openings no one had taken.
Then two weeks of concept sprints. I pushed the team past seventy low-fidelity mobile concepts and held them to one level of fidelity and one visual language, so the set could be tested as a single body of work instead of a patchwork. Each concept was a hypothesis, so we put them in front of real users to find out which ones held.
The interviews decided the MVP. I organized the feedback so the team could synthesize quickly, and the features and concepts that carried the most weight fell out of it. I cut the set to fifty-six screens, built the leading flows into a prototype, and ran a formal usability round with fifteen participants across the core money-movement scenarios. Those findings went back into the design before we built the vision deck and co-presented it to senior leadership.
AI coordinates the rails people already use
Research brought us back to three priorities: speed, cost, and certainty. The design direction used AI to coordinate the payment rails people already use, rather than ask them to learn the infrastructure underneath.
Bring the next decision into view
The proposed experience would learn how someone pays and surface what is owed and available to act on. The value was in helping people see the next useful step, not simply adding a chat interface to existing screens.
Recommend, then let the person decide
AI would suggest the next move rather than act on its own. I used that distinction to keep the concepts focused: assistance should make a payment decision clearer while leaving control with the person moving the money.
Follow the money through to arrival
The experience carried status from scheduled to delivered. Knowing that a payment arrived was part of the core experience, not an afterthought once someone pressed send.
A research-grounded direction for what to build
The consumer vision became an MVP direction leadership backed. This was strategy and concept validation, not a shipped AI payment service. The outcome was a research-grounded direction for what to build and how to keep it understandable.
