Work/Fleet Card

Fleet Card.

A real-time spend-controlled card for fleets. New product line, new revenue, designed from a single index card.

Project metadata

Client: Edenred
Role: Senior Product Designer
Team: 1 PM · 5 Eng · 1 Designer
Shipped: Q2 2022
Type: Strategy0→1
Result: 200+ businesses converted

Friction

— Pain points for the customer

↗There were many strong product ideas but no clear place to start.
↗The company needed to launch a brand-new product line on an aggressive timeline.

My Role

— What I owned during the process

↗Led the design of a new platform from 0→1.
↗Built a scalable design system from scratch and aligned with engineering on implementation.

Outcome

— What I shipped

↗Launched Edenred's first Fleet Card platform in the U.S.
↗Helped secure the VISA partnership.
↗Created a solid foundation for future products in the category.

The TL;DR.

Summary

I designed and launched Edenred USA's first Fleet Card platform, taking it from concept to live in under a year. Built with VISA, it enabled real-time spending controls, opened a new market for the company, and converted more than 200 businesses.

New market. Very short runway.

The Problem

Edenred wanted to enter a fleet card market already dominated by a few strong incumbents. To do it well, the team needed an entirely new platform, a workable MVP, and support for a strategic VISA partnership.

The deadline was aggressive enough that every choice about scope and architecture had long-term consequences.

Key Decision

"Where do we draw the line for what MVP means?"

Ambition was not the problem. The hard part was deciding what had to exist for launch versus what could come after customers were already spending in the product.

MVP meant getting cards into customers hands with real-time controls in place, then building outward from that foundation instead of waiting for a perfect first release.

System first. Product second.

The Solution

With a fixed launch date and a partner watching, the only way to move fast later was to spend the first weeks on things that would not need redoing.

Build the design system early

I defined the visual system and component direction before the screen work ramped up so customer and admin experiences would stay aligned.

Create a flexible shell

The platform architecture needed to support current workflows and future expansion, so the shell was designed to hold more than just the MVP.

Work in short cycles

Features were scoped into roughly two-week chunks and validated continuously with internal stakeholders and VISA to reduce risk as the deadline approached.

Concept to launch. Under a year.

The Process

Exploration

Card status was the whole product, so the wireframes started there: Active, Frozen, Pending, Fraud, all legible in one table.

Cards list with status states

Every card state — active, frozen, pending, fraud — laid out at once, to check the table could carry all of them without turning into noise.

Step 01

Define the architecture

I started by shaping the platform shell and the information architecture so the rest of the work had a scalable place to land.

Step 02

Design for customers and admins

The customer side needed real-time spending controls, but the admin side had to be just as scalable and intuitive for the product to hold up in practice.

Step 03

Stay close to implementation

Strong engineering alignment and ongoing design oversight were essential because front-end decisions had a huge effect on whether the product stayed coherent.

Shipped

Under a year from index card to live product.

Every transaction, with context

Merchant, fuel product, gallons, PIN, odometer, and exception flags on a single row — what a fleet manager needs to spot a bad purchase without opening it.

Card states, all of them

Active, frozen, canceled, pending, lost, stolen, fraud. The status column had to carry the entire lifecycle and stay scannable.

Real-time spend controls

Limits per card by transaction, day, week, and month, each showing what has already been spent against it. This was the product.

Exceptions, ranked

Which policy breaches happen most often, over what range, and on which cards — turning a pile of declines into something a manager can act on.

What worked. What got tricky.

Reflection

Tough spots

Some post-MVP features turned out to matter more to customers than expected.
Engineering handoffs were painful because front-end expertise was uneven.
MVP discipline had to be defended constantly against understandable ambition.

What went right

The team moved fast and still shipped on time.
The project set a new internal standard for how a new product could be built and launched.
Starting with a code-aware design system reduced rework later on.

Edenred - 2023

Commuter Benefits

Personal Project - 2026

RemoteTechRoles