AI Financial Copilot — Next Moves
The problem
Most financial apps show you what you have. They don't tell you what to do next. This project explores shifting financial products from passive dashboards into decision-making systems.
“Where should AI make decisions, and where should it not?”
The approach
Contributing to RRSP, paying down debt, allocating savings — the moments where users are most uncertain and most in need of direction.
Readiness score + urgency flag, derived from user inputs: income, accounts, contribution room, goals, deadlines. Transparent and auditable — not a black box.
Ranked actions are sent to Claude, which generates one personalized sentence per recommendation. Example: "You got a refund last year and haven't maxed your RRSP — this unused room is costing you in tax drag."
As users complete actions, scores update in real time — turning the tool from a recommendation engine into a behavior engine.
AI for articulation, not decision-making. Financial decisions require trust; black-box recommendations reduce confidence. Deterministic logic handles the actual reasoning — AI adds context, motivation, and personalized communication on top of it.
“Most people don't need more financial information. They need clarity on what to do first.”
Live demo → ws-ai-challenge.vercel.app← All work