AI-Powered Corporate Card Management Platform

Initially, users had one simple option — a virtual credit card — so the flow was straightforward because there was only one possible path. With new card offerings, users needed the flexibility to create virtual credit, physical credit, virtual debit, and physical debit cards, each with different behaviors and requirements: physical cards needed shipping information, debit cards needed account connectivity, and finance teams often needed to issue cards for themselves or across a whole team.

The challenge was never just adding more options. It was: how do we introduce flexibility without overwhelming users with complexity — while staying scalable for whatever card type came next. Introducing physical cards wasn’t just a UX improvement — it was a business initiative. Enabling physical card usage meant Zeni could increase real-world spending through company-issued cards, driving greater adoption of Zeni’s financial ecosystem, and positioning the platform as a more complete card management solution for growing businesses. That meant the design had to balance business growth goals, operational flexibility, user clarity, technical feasibility, and scalability for future expansion, all at once.

Before jumping into screens, I focused on how introducing multiple card types would affect the overall experience — how users could tell card types apart, how to avoid decision fatigue from too many options, how physical and virtual experiences should differ, and how the system could keep scaling for card types that didn’t exist yet. One insight shaped everything downstream: simply adding more forms or dropdowns would create confusion. Users needed a guided experience, not additional complexity.

The final experience introduced a unified, scalable card-creation journey. Visual card previews replaced a plain list so users could recognize and choose between card types at a glance; the creation flow itself adapted to what a user selected instead of front-loading every possible field.

• Clear card selection — visual previews replaced a plain list, reducing ambiguity between options.

• Guided creation flow — physical cards prompted shipping info only when relevant; debit cards asked for account-linking only when needed.

• Multi-user card assignment — finance teams can now assign cards to themselves or other employees more efficiently.

• Improved activation journey — redesigned onboarding and activation created better visibility and less friction after issuance.

How might we design an AI experience for financial operations where trust, accuracy, and human oversight matter? Unlike consumer AI products, fintech workflows can’t rely on assumptions — financial actions require precision, explainability, clear confirmation, and operational trust. This wasn’t only designing a chatbot. It was designing an AI operational agent capable of understanding intent, asking intelligent follow-up questions, and completing card management tasks safely.

Instead of navigating complex menus, users simply describe what they want — for example, “create cards based on our top vendors.” The AI agent interprets intent, analyzes financial context, and initiates an intelligent workflow. Rather than executing blindly, the system reasons through the task and asks contextual questions before taking any action — a balance between automation and accuracy, not automation instead of accuracy.