Helping people build financial confidence with AI

Problem

What we heard

The founder led the user research, and I designed from what it found. People kept running into the same barriers: too little financial education, no clear view of their money, conflicting advice, fear of risk, no saving habits and low confidence. These barriers shaped the whole product structure.

• Knowledge gaps. “I feel like I’m losing an opportunity. I don’t have the information or the security to do something different.”
• No clear picture. “I don’t understand why we make so much money, yet still struggle with debt. Where is it going?”
• Too many voices. “I was getting fragmented information from different people, so it was hard to know what to do.”
• Low confidence. “I didn’t feel confident doing it. It takes a lot of effort to look at things.”

Every barrier came back to confidence. So confidence became the product.

Product structure

Structuring the product

From the research, I identified what people were looking for and what their current tools were missing. Then I worked out how those needs could fit together logically in one product, with one simple model: help people understand money, manage it, and act with confidence. In a brainstorm session with the team, we defined the app’s goals, areas of engagement and target personas.

The map below shows the full journey I defined — from first questionnaire to daily use — so every flow had a clear job before any screen was designed.

Understand, manage, act. One simple model gave every flow a clear job.

Onboarding

Understanding each user from the start

First impressions decide whether people stay. I designed a welcoming sign-up and a step-by-step questionnaire that asks only what’s needed to personalise the app: goals, priorities and where each person is starting from.

Key decisions

• Ask only what changes the experience, so every question earns its place
• One question per screen, to keep effort low and progress visible
• Goals and priorities first, so personalisation starts on day one

Fewer questions, better answers.

The two prototypes below play the welcome screens and the personalisation questions. Tap either one to take over.

Education

Learning in bite-size steps

Financial confidence starts with understanding. I designed the learning experience to explain financial products and money habits in small, digestible pieces.

Friendly language, rooted in real life. Every piece of knowledge links money to everyday situations. A savings account is “your go-to handbag”, a workplace pension is “a duet with your employer”. Familiar comparisons turn a daunting topic into something approachable.

Two formats, two jobs. The Financial Products library explains what’s out there: what each product is, how it can grow and how risky it is. Bestie’s Top Tips answers the questions people actually ask, like how much to save or how to negotiate a pay rise.

Small steps feel achievable. Achievable steps become habits.

Both prototypes below are clickable. They play through on their own; tap anywhere to take over and explore.

Game

Learning money through play

Practising makes real decisions less frightening. I designed a game where people try real-life money situations in a safe, low-stakes way. Abstract concepts become decisions with consequences they can feel, reinforcing the learning elsewhere in the app.

Key decisions

• Base scenarios on real financial situations, not abstract quizzes
• Show the outcome of each choice straight away
• A playful visual language, so a serious topic feels safe to try

Play gave people permission to get things wrong.

Tracking

Seeing the whole picture and where it’s going

A goal feels real once you can see it. Users connect their accounts for a real-time view of their money and set custom spending ratios across essentials, non-essentials and future investments.

Key decisions

• Show the future as well as the present: a projection that reacts to every choice
• Let people set their own spending ratios instead of imposing a budget
• Categorise transactions automatically, so tracking takes no effort

A goal you can see growing is a goal you keep.

The prototype below is clickable. Each account becomes a tree sized by the money in it. It plays through on its own; tap anywhere to take over and explore.

AI assistant

Turning data into plain-language guidance

Generic finance advice doesn’t build confidence. I designed a conversational AI assistant that reads a user’s situation and answers with personalised, plain-language guidance.

The voice was designed as part of the brand, to match the app’s bold, playful visual style. We trained it to sometimes sound like Cher: confident, glamorous and a little cheeky, but never preachy.

Key decisions

• A warm, collaborative voice over an instructive one, because research showed it builds trust
• Suggest next steps, but leave the choice with the user
• One consistent tone across the whole app, not only in the chat

An assistant earns trust through how it speaks, as much as what it knows.

The animation below plays a typical conversation with Money Bestie, from a real money question to personalised next steps.

Visual identity & design system

A system built for visual freedom

Finance can feel intimidating, so WInvest needed to feel the opposite. I gave it a bright, energetic identity: vibrant colours, friendly type and a warm illustrative style, designed together with the playful, Cher-inspired tone of voice.

The system was deliberately loose. Instead of strict brand rules, it fixed only the foundations: type, spacing, core components and their states. Everything above that, especially colour, stayed open, so screens could feel fresh and expressive while the product still behaved consistently.

I built it alongside the product and documented it, so design and engineering shared one source of truth without slowing a small team down.

Key decisions

• Strict on the foundations, free on expression: type, spacing and components fixed; colour and illustration open
• Colour as a source of energy, not a rule to follow
• Visuals and tone of voice designed together, so the app looks the way it speaks
• The system built alongside the product, so it grew with every feature

Structure where it matters, freedom everywhere else.

Outcome

A complete product, shipped

I took WInvest from first concept to a launched iOS app. As founding designer I owned the full user experience: onboarding, the AI assistant, tracking and projection, learning, the game, the visual identity and the design system.

It turns an intimidating problem, managing money across budgeting, spending and investing, into something approachable, built on a system designed to scale. WInvest is live on the App Store.

What I’d measure next

• Activation through the first goal set
• Week-four retention
• Engagement with the AI assistant and learning content

Key takeaways

• Every barrier came back to confidence, so every feature was designed to build it.
• For an intimidating topic, tone and approachability do as much as features. The supportive AI voice, validated in research, shaped every interaction.
• Making the future tangible drives action. The net-worth projection turned an abstract goal into something people could see and adjust.
• A deliberately loose system kept the app coherent and expressive without slowing delivery.