Led UX strategy and product definition for Benjamin, a 0→1 AI-powered personal finance platform targeting the US market — translating ambiguous business opportunities into a concrete product vision, user model, and interaction framework, currently in executive review.
Role
Product Designer
Team
1 Product Manager
2 Designers
5 Engineers
Timeline
August 2025 - Present
Tools
Figma, Miro, Google Docs, Slack, and etc.
Overview
Benjamin is a 0→1 AI-powered personal finance platform being developed at Hanwha AI Center — designed to give everyone access to a personalized CFO in their pocket. I led the end-to-end UX strategy and product definition, from initial research through executive presentation, translating ambiguous business opportunities into a concrete product vision, interaction framework, and MVP concept.
The concept received full executive approval and is now moving into development.
The Problem
Most Americans know they should manage their finances better. But knowing isn't the problem.
65% of Americans have no financial plan. Existing tools like Credit Karma, Rocket Money, and Cleo offer analysis or coaching — but none connect scattered financial data into a single agent that can understand your situation, tell you what to do first, and actually execute on your behalf.
The real gap isn't information. It's the space between insight and action — and the trust required to bridge it.
Research
Phase 1 – Early Exploration
We started with a broader hypothesis: a budgeting app targeting Gen Z. To validate it, I led mixed-method research including 25+ surveys and 8 user interviews, synthesizing findings across financial behaviors, attitudes toward AI, and unmet needs.
What we found
Users fell into two distinct groups: self-driven (uses Excel, builds their own portfolio) and not confident (wants guidance, scared to make financial decisions)
Both groups shared one thing: they don't know what to do first
Trust was a significant barrier — users were skeptical of handing financial decisions to an AI they didn't understand
Budgeting features had demand, but complex setup caused immediate drop-off
The most consistent signal: "Simple UI is the best" — users wanted a clear starting point, not a comprehensive dashboard
Phase 2 – Reframing the Problem
The research revealed that a budgeting app alone wouldn't solve the real problem. The issue wasn't that people couldn't track spending — it was that they couldn't confidently act on what they knew.
This led to a product pivot: from a Gen Z budgeting tool to a Personalized AI CFO — an agent that doesn't just show you your finances, but tells you what matters right now and helps you do something about it.
"People are highly interested in building wealth but don't have the financial knowledge — and are scared to make any financial decisions."
The opportunity
The AI/LLM market is growing at +111% CAGR. Financial AI agents are still early — most players focus on a single layer (reporting, coaching, or execution) but none connect all three.
Benjamin targets the gap: behavior change through intelligent, executable recommendations — positioned between emotional chat coaches like Cleo and rational reporting tools like Credit Karma.
Product Direction
Core Value Proposition: Your Pocket AI CFO
Three pillars shaped all product and UX decisions:
Your
Deeply personalized to your actual financial situation, not generic advice
Surfacing what's relevant right now, not information overload
CFO
Not just analysis. Forecasts, prepares, recommends, and executes on your behalf
Solving the Trust Problem
Trust was the hardest problem to crack — and our biggest differentiator.
Most AI finance tools ask users to trust a black box. Benjamin solves this differently: AI handles the complexity, humans handle the commitment.
The flow works in three stages:
AI simplifies
Complex bills, insurance policies, phone plans are analyzed and summarized in plain language. Benjamin tells you what you're actually paying for and what you could save.
Broker connects
For high-stakes decisions (insurance, loans, phone plans), real licensed brokers step in. Human expertise where it matters most.
Comparison made easy
After broker consultation, users get a clear, visual side-by-side comparison to make confident decisions without feeling overwhelmed.
This hybrid model — AI efficiency + human trust — is what separates Benjamin from pure AI tools and expensive traditional advisors.
Product Principles
Defined four principles that guided all UX decisions:
Clarity over completeness
Show the one thing that matters most right now, not everything at once
Action over information
Every insight leads to a concrete next step
Trust through transparency
Explain the "why" behind every recommendation
Friendly but rational voice
Encouraging, never shame-based. Accurate, never overpromising
Key Screens
Home Dashboard
Top priorities surfaced immediately with total potential savings. Financial overview at a glance — not a data dump, but a prioritized action list.
Bill Detail + AI Recommendation
Contextual CFO advice tied to real bill data. Benjamin explains what's changing, what it means for your budget, and what to do about it.
Conversational AI Interface
Ask Benjamin anything. Proactive financial guidance, with budget impact previews tied to real numbers.
Document Upload
Upload bills, insurance policies, and financial documents once. Benjamin reviews, categorizes, and surfaces what needs attention.
Process
The product direction emerged from a series of cross-functional ideation workshops — mapping user characters and needs, defining core value props, and stress-testing the AI vs. CFO interaction model before committing to a design direction.
Status & Next Steps
The concept received full executive approval and is now moving into development.
If approved, my immediate next priorities would be:
Validate the "Top Priorities" model with usability testing — does it actually reduce decision fatigue?
Test AI voice and tone — where does encouragement tip into overpromising?
Define the onboarding flow — goal-first vs. data-first entry point
Design the broker handoff experience — the critical trust moment in the product