TheAI
FintechAI Product Design
Designing a trust-first AI recommendation flow for financial decisions.
ROLE
Product Design Intern
TEAM (4)
Product DesignerX2
Product Manager
Research&Marketing
Engineerx2
TIMELINE
05/2025-09/2025
TOOL
Product Context
FROM CHATBOT TO FINANCIAL AGENT
TheAI is an AI-powered financial guidance product that helps users make complex personal finance decisions through conversation.
When I joined, an early version of the product was already in place, adapted from WeFire's existing chatbot experience. It could answer financial questions and collect user information, but the interaction model was still relatively linear and the product experience was largely centered around a single conversation..
My Contribution
I worked across interaction design, visual design, prototyping, and product validation to evolve the early chatbot experience into a more complete AI financial product.
Quick Summary
PROBLEM
Required inputs interrupted the flow
Product-team research suggested that 30% of the tested users found it frustrated and time-consuming when a conversational AI required personal financial information before the conversation began.
BUSINESS GOAL
More personalized AI advice
The product team believed richer financial context was necessary for TheAI to provide more useful recommendations and stand out from generic AI chat tools.
DESIGN SOLUTION
Flexible context collection
Instead of treating missing information as a blocker, I redesigned the flow so users could provide full context, continue with partial context, or preview a complete answer with demo data.
OUTCOME
50% less time spent on required input
Follow-up testing showed an approximately 50% reduction in time spent completing the previously required financial information step.
The Product Tension
The product needed context. Users needed value sooner.
When I joined, users had to complete relevant financial details after asking a question before receiving a personalized answer. The logic supported better recommendations, but testing showed that the required step could create friction before users had seen enough value.
WHY IT EXISTED
Structured financial context helped TheAI provide more specific recommendations.
WHAT USERS FELT
3 of ~10 participants described the required information step as long or unusual.
The Tension
The issue was not whether to collect context, but when and how much to ask for.
What The Evidence revealed
I looked at both user behavior and existing interaction patterns to understand where the friction came from and whether personalization really needed to begin with a complete profile
USER EVIDENCE
Users understood why personalization mattered, but the value came too late
Signal 01:
5 / 10 participants
described the required information step as long, unusual, or too much before seeing value.
Signal 02:
4 / 10 participants
expected more beyond the conversation, describing the experience as useful but still too similar to a general AI chatbot.
COMPETITIVE EVIDENCE
Other products often built context progressively, asking for additional information after users understood why it was useful.
Across these patterns, context was collected in smaller, more contextual moments rather than as one large upfront task. This suggested that TheAI didn't need less financial context, it needed a better moment to ask for it.
This raised a core product question:
How might we help TheAI give useful advice with incomplete context, while still encouraging users to provide more information when it matters?
Design Decision
01
Designing for incomplete context
The goal was not to remove financial context, but to stop missing information from blocking the conversation. I redesigned the required profile step into a more flexible flow, allowing users to move forward with partial context and add more information when it became relevant.
Proposal A —Selected For Final Prototype
Progressive Personalization
User Flow
Ask a question → Initial answer → Add some / all information → Continue conversation → Contextual callback when more is needed
Users are no longer required to complete a full profile before receiving value. They can provide as much or as little information as they want, while TheAI uses contextual callbacks later to request missing details only when they become relevant to the recommendation.
Rationale
Turn profile completion from a prerequisite into a progressive part of the conversation.
Pro
Reduces upfront effort and gives users more control over how much information they provide.
Con
Early answers may be less precise until enough context is collected.
Proposal B —Rejected After discussion
Let users preview TheAI before completing their financial profile.
User Flow
Ask a question → Demo answer or Complete financial profile → Personalized answer
Users can see a representative answer without completing their financial information first. When they want a personalized recommendation, they are prompted to finish their profile.
Rationale
Show the product’s value before asking users to commit to setup.
Pro
Delivers an immediate preview while keeping the existing personalization flow largely unchanged.
Con
The first answer isn’t truly personalized, and users still face the same profile setup later.
02
Conversation context into a persistent financial experience
While progressive personalization improved the conversation, users still had limited visibility into what TheAI knew about them beyond the chat. I designed a Financial Profile to make that context visible, editable, and reusable across future recommendations.
1.
Visible context
Users can see what TheAI already knows about their financial situation.
2.
Editable information
Financial details can be reviewed and updated outside the conversation.
3.
Reusable across conversations
Profile updates can improve future recommendations without asking users to repeat themselves.
Deliver
From question to personalized recommendation
The final flow allowed users to start with a broad financial question, add only the details they were comfortable sharing, receive an answer with current information, and improve the recommendation later by adding more context.
Visual Design
I built the visual system with another product designer around clarity and restraint, using a calm dark palette, consistent components, and clear hierarchy to make complex financial information easier to scan and understand. The goal was to keep theAI feeling trustworthy and focused without making the experience feel overly technical.
Impact
From a rebranded chatbot to a more complete AI financial product.
The work helped clarify TheAI’s standalone product direction, shifting the experience toward earlier value, progressive context collection, and a persistent financial profile. The final interaction model was also reviewed with adjacent product teams, helping validate the overall direction and refine smaller issues around hierarchy and flow.
~50% less setup time
Progressive collection reduced time spent in the information-entry portion of the tested flow.
Cross-team validation
The core interaction model held up in reviews with adjacent product teams, with smaller flow and hierarchy refinements made afterward.
Reusable product system
Conversation states, Financial Profile patterns, and shared components created a stronger foundation for future product work.
