🥇 1st Place · TechTO Hackathon × Tangerine

Making money
make sense.

An AI finance manager that meets newcomers, new grads, and first-time mortgage holders where they are. Clementine translats Canada's confusing financial system into plain language.

Role
UX Research & Design
Timeline
10 Hours
Team
3 People
Year
May 2026
Clementine dashboard showing a personalized paycheque breakdown

Clementine's home dashboard. Every paycheque, finally explained.

The Spark

It started with my mom.

When my family immigrated to Canada, I watched my mom struggle to understand a financial system that assumed she already knew the rules. Years later, as a new grad entering the workforce, I hit the same wall why was part of my paycheque just gone? What were these "mysterious" bank fees? What on earth is an RRSP?

During our hackathon brainstorm, my teammates realized they'd all asked the same questions. The financial system isn't just complex, it's quietly exclusionary to anyone who didn't grow up inside it. So we set out to build something that closes that gap.

"We believe AI should fill a real gap in people's lives, not just exist for the sake of it."
The Problem

Financial literacy tools assume you already speak the language.

Existing banking apps show you numbers, not meaning. For someone new to Canada or new to the workforce, that's not help, it's homework. We anchored our design around three questions real users actually ask:

"Where did my paycheque go?"

Deductions like income tax, CPP, and EI appear with no explanation of what they are or why they're normal.

"What's this fee for?"

Bank fees feel arbitrary and hidden, with no guidance on how to avoid or reduce them.

"What even is an RRSP?"

Acronyms like RRSP, TFSA, and CPP are everywhere — and nowhere explained in plain language.

The Process

From zero to working product in ten hours.

With the clock running, we moved fast but stayed disciplined — research first, design grounded in a real persona, then build.

01

Research & Synthesis

We mapped the pain points shared across our own newcomer and new-grad experiences, benchmarked existing players to find our differentiation,then plotted every idea on an impact–effort matrix and explicitly marked what not to build (view our impact-effort matrix →)

Competitive analysis from the brainstorm
Our competitive analysis. We compared Clementine to existing financial literacy tools and found our competitive edge.
02

Persona: Meet Angel

We built our design around Angel Chen, a newcomer working her first job, financial-literacy beginner, juggling RRSP decisions, bank fees, and building credit. Every screen had to work for her.

Persona card for Angel Chen showing her profile, financial stats, goals, pain points, knowledge gaps, and AI engagement principles
Angel's persona card. It has her goals, knowledge gaps, and pain points anchored every design decision. We even defined how the AI should speak to her: warm, non-judgmental, never assuming prior knowledge.
03

Design & Prototype

We mapped the full user flow from sign-up and personalization quiz to the generated dashboard, education prompts, and AI-guided actions. We then designed a dashboard that explains rather than just displays.

Complete user flow diagram from sign-up through personalized dashboard, education, and AI-guided actions
The complete user flow. The pink path is the core journey: sign-up → personalization quiz → a dashboard generated around you. Click to view full size.
The Solution

Three ideas that make Clementine click.

01 — EXPLAIN, DON'T JUST DISPLAY

A paycheque that teaches

Instead of a wall of deductions, Clementine breaks down gross pay into plain-language pieces and reassures the user when something is normal for their income and province. Every line has an "explain in plain language" option.

Paycheque breakdown with plain-language explanations
02 — MEMORY THAT ADAPTS

AI that remembers your context

Clementine personalizes itself around what matters to each user. It remembers that Angel is a newcomer from China, new to the work force, a financial-literacy beginner with goals like reducing fees and improving credit. Users stay in control: every memory is transparent, toggleable, and deletable.

Settings screen showing what the AI remembers about the user
03 — GROWS WITH THE USER

Literacy levels, not one-size-fits-all

During onboarding, users set their own financial literacy level, from "just starting out" to "comfortable". Clementine adjusts how much it explains. The microcopy does quiet work here: "Be honest — there's no wrong answer." Designed for people used to feeling embarrassed by what they don't know.

The Outcome

First place — and a problem worth continuing.

Out of a room of strong teams, Clementine took 1st place at the TechTO Hackathon sponsored by Tangerine. More importantly, judges and fellow participants connected with the core idea: technology that meets people where they are.

1st
Place at TechTO × Tangerine
10
Hours from idea to prototype
3
User segments served
Reflection

What I'd carry forward.

Constraints sharpen design

Ten hours forced ruthless prioritization. We designed only what served Angel.

Plain language is a feature

The biggest UX win wasn't visual it was the idea. Translating jargon into reassurance is what made the product feel human.

AI needs intention

We chose to use AI where it genuinely closed a gap. Transparent, deletable memory kept the user in control. We want trust before novelty.

Next: validate with real users

The natural next step is moderated testing with actual newcomers to pressure-test our assumptions about literacy levels and tone.

Grace Dong · UX & Research