-
-
The Financial Autopsy - see what doing nothing has cost you
-
Fill out your financial profile in under 2 minutes
-
Full inaction breakdown -bento grid with damage chart, time thief, counterfactual you, and recovery speed
-
AI-powered regret stories and recovery roadmap personalized to your location
Inspiration
I kept seeing people around me stuck in the same pattern. Smart people, working hard, but quietly losing thousands of dollars every year just by not doing anything. Not making bad decisions. Just not making decisions at all. Nobody had built a tool that put an exact number on that. So I did.
What it does
WillSpend runs a compound simulation across every passive financial decision you've made. The salary you never negotiated, the savings account you never switched, the investments you kept putting off, the subscriptions you forgot about, the debt you never refinanced. It adds it all up with real compound math and shows you the total damage. Then an AI advisor powered by LLaMA 3.3 70B translates that number into something real. What it would have bought you, and what you can do about it this week.
It supports both US and India users. US users get a 401k match leak calculator. India users get a SIP delay cost calculator. Everything updates live as you type via a sticky ticker at the bottom of the screen.
How I built it
Backend is Python and FastAPI handling the simulation math. Each inaction category has its own compound formula. Monthly compounding for savings and investments, future value calculations for missed investing, interest differential for unrefinanced debt. The AI layer runs on Groq's free tier using LLaMA 3.3 70B, which is fast enough to feel instant. Frontend is vanilla HTML, CSS, and JavaScript with Chart.js for the damage breakdown chart and jsPDF for client-side PDF generation. Deployed on Render and Netlify.
Challenges I ran into
Render's free tier spins down after inactivity which means a 50 second cold start. Users were hitting submit and seeing nothing happen, assuming the app was broken. I built a custom cold start overlay with cycling status messages and a progress bar that fills over 60 seconds. It completely changed the perception. Users now wait because they feel like something is happening.
The other challenge was making the AI output feel personal and not generic. The key was passing the user's city and country to the prompt and asking for location specific regret stories. "That $40k would have been a down payment in Austin" hits differently than "you lost $40k."
Accomplishments that I'm proud of
Built the entire thing solo in under 5 days. The live loss ticker was a late addition that ended up being the most engaging part of the app. Watching the number climb as you fill out the form creates a visceral reaction before you even submit.
What I learned
The math is the easy part. The hard part is making data feel personal. Numbers alone don't create urgency. Context does.
Social Impact
Financial inaction costs the average person tens of thousands of dollars over their lifetime. WillSpend makes that invisible damage visible. It helps people in the US and India understand the real cost of passive financial decisions and take action before it's too late. Most useful for first generation earners and young professionals who never had anyone explain this stuff to them.
What's next for WillSpend
Adding more inaction categories. High fee mutual funds vs index funds, lifestyle creep tracking, and a forward simulator showing what happens if you start acting today.
Log in or sign up for Devpost to join the conversation.