Personal Finance Dashboard with Auto-Recommendation Engine
Full-stack app for tracking expenses, loans, and savings — with a math engine that recommends monthly allocations.
Outcome
A live, authenticated full-stack web app that ingests income, fixed expenses, credit cards, and loans, then runs a recommendation engine to suggest optimal monthly payment allocations. Features a strategy comparison banner that shows the trade-off between paying loans in parallel vs. avalanche (e.g. avalanche saves 16 months and $4,443 in interest, but uncles wait 18 months for first payment). Used daily for real financial decisions on a $44K student loan and personal cash flow.
Motivation
Personal finance apps default to interest-bearing optimization (always pay the highest-interest loan first). For someone with values around Islamic finance (avoiding riba) and family loans, that math is wrong — uncle loans at 0% would never get paid back in pure-math mode. I needed a tool that respected both math and human values, with an explicit 'force include' flag per loan and overridable recommendations.
What I did
- →Architected a Next.js + Supabase stack with RLS-secured data per user
- →Designed a month-by-month loan simulation engine handling compound interest, moratoriums, and zero-interest edge cases
- →Built a recommendation algorithm that respects user-flagged loans before sending surplus to highest-interest debts
- →Implemented a strategy comparison engine: parallel vs. avalanche payoff with full interest accounting
- →Verified all financial math against console-logged real numbers before deploying
Technical detail
The math engine is the centerpiece. `simulateLoanPayoff` runs month-by-month: applies interest at the monthly rate, subtracts EMI + extra payment, detects edge cases like 'payment doesn't cover interest' and reports 'balance grows forever' rather than silently looping. `simulateBudgetMode` handles cascade — when one loan finishes, its monthly budget slice automatically reallocates to the next priority. The recommendation algorithm pulls force-included loans first (their balance divided by target months), then routes the surplus to the highest-interest interest-bearing loan. All amounts handled in USD using a user-controlled INR/USD rate so the loan in rupees is comparable to the loans in dollars.