An AI-powered financial clarity layer built to help everyday users understand where their money goes and feel genuinely in control without adding more work to their day.

My Role: Lead Product Designer

Product Type: FinTech

Team Members: Laura, Jisoo, Jennifer

01 OVERVIEW

What is this project?

Reveal is an AI-powered financial wellness experience designed to help users better understand, track, and take control of their spending habits.

Built as a B2B plugin for the AMEX app, the project focuses on reducing financial stress by transforming complex transaction data into clear, actionable insights.

Through in-depth user research, Reveal uncovered how users often rely on mental math, avoid checking statements, and struggle with unclear transaction details.

The solution introduces intelligent spending summaries, proactive money pulse notifications, gamified saving challenges, and an AI assistant that helps users make more confident financial decisions.

"The problem isn't your spending. It's the lack of visibility." The insight that shaped every design decision in this project.

My Role

I served as the Lead Product Designer on a team of four, overseeing project timelines, sprint planning, stakeholder communication, task prioritization, and cross-functional coordination throughout the project lifecycle.

Alongside managing key product and project management responsibilities, I led the end-to-end design process conducting user research, defining user flows, creating wireframes, designing high-fidelity UI screens, and ensuring design consistency across multiple features and touch points.


02 PROBLEM

What problem are we solving?

User research revealed that people weren’t struggling with spending alone they were struggling with understanding their financial activity clearly and confidently. Users often relied on memory, invoices, notifications, or external apps to verify transactions and track expenses.

💵

Users lack a clear, organized view of their spending, making it hard to track monthly expenses, identify patterns, and stay in control of their credit usage.

💳

Users struggle to quickly identify and verify unfamiliar charges, often relying on manual investigation through transactions, receipts, and notifications to confirm legitimacy.

🧾

Users struggle with unclear and inconsistent transaction details, making it difficult to recognize purchases, verify charges, and confidently track their spending.

RESEARCH

What we heard when we asked people about money

I conducted user interviews to understand how people emotionally and practically respond to credit card bills. We weren't just asking what features they wanted — we were listening for what made them feel anxious, confused, or in control.

The research surfaced something more nuanced than expected: people weren't struggling with where to find financial information. They were struggling to contextualize it. A number without context is just noise.

😰

Anxiety over ambiguity

Users felt stressed by merchant names they didn't recognize — not knowing if a charge was legitimate or fraudulent created a background hum of financial anxiety.

📊

Charts without meaning

Users looked at charts but couldn't act on them. Data was presented — never explained. They could see what they spent, not why it mattered.

🔁

Subscription blindness

Almost every participant had at least one subscription they were paying for but no longer using — and didn't realize it until asked directly.

🤝

Wanting a conversation

Several users said they wished they could "just ask someone" what a transaction was — a direct signal for an AI-native solution like Reggie.

— 02 — MARKET

The market gives users data. Reveal gives them clarity.

We analyzed six major platforms: Apple Wallet, American Express, Chase, Rocket Money, Capital One, and Monarch Money. Each had real strengths — but three gaps appeared across all of them.

Every tool waits for you to notice a problem. None tell you something changed before you discover it yourself, often too late.

No proactive insights

Subscriptions are afterthoughts

Most apps list recurring charges but none alert you before they hit — and none make cancellation feel easy or obvious.

Transaction data is presented clinically — no context, no tone, no explanation of what the numbers mean for your specific situation.

No human layer

Reveal fills all three gaps: proactive insights, a human conversational layer via Reggie, and full subscription control — inside AMEX.

Where Reveal leads

— 03 — OPPORTUNITY

Defining the design challenge

HOW MIGHT WE

Help everyday users better understand their spending, catch hidden charges, and make smarter financial choices — without adding more work to their daily routine?

Before building, we paused to validate. We ran concept tests with lightweight prototypes to see which features landed and which created friction.

WHAT CREATED FRICTION

Manual tagging. Charts users had to interpret. Notifications that felt like scolding. Overwhelming dashboards. Anything that added more work instead of reducing it.

WHAT USERS LOVED

One-swipe monthly summaries that explain changes. Subscription alerts before charges hit. Gentle reassurances: "You're still within your usual range." Passive, effortless insight.

This led to our clearest product principle: users don't want to analyze their money — they want to understand it at a glance, feel in control, and get on with their day.

We chose to build inside the AMEX app rather than as a standalone product. Users don't want another app. They want their existing tools to get smarter. Reveal adds a human layer on top of AMEX — no migration, no extra login, no mental overhead.

— 04 — DESIGN

From user flows to a system that feels calm

Our wireframe process moved through three fidelity levels — rough sketches, mid-fi flows, high-fi prototype. At each stage we tested with users and adjusted. The most significant pivots came from watching where people hesitated.

Primary user journey from landing to financial clarity.

Mid-Fidelity Wireframes showcasing insights dashboard, transaction detail, subscription manager, Reggie entry point.

Logos

Chat Bubbles

Buttons

Color Palette

Navigation Bar

Icons

Emojis

Health Tracker

Charts

Logos

Trend indicators

Design system for Reveal

— THE SOLUTION

Bringing Clarity to Everyday Finances

Managing finances is tough with unpredictable spending and hidden subscriptions. This AI-powered solution with a subscription tracker helps users monitor spending, manage recurring payments, and get personalized recommendations for better financial control.


Meet Reggie, Reveal’s AI bot helping user understand transaction details and more…


Financial Insight with multiple widgets giving detailed information into user spending

Detailed insights to understand spending spikes and easy dispute of transactions


— 05 — USABILTY TESTING

The chatbot placement that changed everything

We tested three flows with real users: exploring spending insights, managing subscriptions, and interacting with Reggie. The feedback reshaped all three — but the Reggie placement was the biggest learning of the entire project.

V1 — INSIGHT DASHBOARD

Charts were small and dense. Users struggled to understand what changed or why it mattered. Key metrics were buried. The primary view had too much competing for attention.

V1 — REGGIE (FAB)

Floating action button with a bot icon. 60% recognition. 40% had no idea what it did. Users weren't confused about AI, they were confused about context.

V1 — SUBSCRIPTION

The word "block" confused users. They couldn't tell the difference between blocking and cancelling. Post-blocking steps were unclear, creating hesitation and distrust.

V2 — INSIGHT DASHBOARD

Simplified to one key metric per module with a drill-down layer for detail. Added descriptive language ("Up 22% this month — mostly from dining") alongside the chart, not instead of it.

V2 — REGGIE (CONTEXTUAL)

Moved below transaction detail with a plain-language prompt. Near-universal recognition. No redesign needed, just the right moment, the right place.

V2 — SUBSCRIPTION

Replaced "block" with clear action language. Added an explanation of what happens post-action. Early communication of limitations boosted trust and completion rates.

In V1, Reggie was a floating action button — 40% of users had no idea what it did. We moved it below the transaction detail with the prompt "Not sure what this charge is? Ask Reggie." Recognition went near-universal. Same feature. Different placement. Completely different result.

— 06 — FINAL DESIGN

Reveal — a human layer for credit card experiences

Reveal isn't just an insight engine. It's the layer that makes the complex feel calm — sitting inside the AMEX app, available exactly when users need it, invisible when they don't.

100%

Reggie recognition after contextual placement fix

3

Flows designed & tested end-to-end

V1 → V2

Full iteration driven entirely by test data

01

02

03

04

— 07 — FINAL DESIGN

Reflections

Emotional safety is a design constraint. When we tested a spending alert that said "Your dining spend jumped 40% this month," users felt judged — not informed. We rewrote every piece of data communication to be descriptive rather than evaluative. That shift from "you overspent" to "this is what changed" is what made the product feel supportive rather than punishing.

Co-leading taught me that product decisions are negotiation. Managing timelines, pushing back on scope creep, and keeping a four-person team aligned while each of us owned different flows — that stretched me beyond design craft. I learned to make tradeoff decisions with incomplete information and communicate them clearly.

Feature placement is half the design. Moving Reggie from a FAB to a contextual prompt below the transaction detail improved discoverability dramatically with zero visual redesign. Where a feature lives shapes whether users reach for it.

Systems thinking isn't optional in financial design. Every screen in Reveal exists in relationship to every other screen. Designing one widget without thinking about how a user got there — and where they'll go next — produced confusing flows in V1. Building a proper design system early would have saved us from inconsistencies we had to fix in iteration.

Next Project

Made with ❤️ and lots of iced lattes