
Production-Grade SaaS with AI Insights
A comprehensive, AI-driven personal finance (PFM) application with a pro-tier subscription model, featuring automated expense tracking via OCR, budget management, and intelligent financial insights.
Defining the core problem and identified pain points that necessitated this technical intervention.
Users struggle with the time-consuming manual data entry required for accurate budgeting and lack clear, actionable insights into their spending habits. This leads to budget overruns, missed bills, and a failure to meet long-term financial goals.
The architectural and implementation strategy developed to resolve the challenge.
Full CRUD for Income, Expenses, and Bills. Every transaction atomically updates a central 'balances_collection' and a 'monthly_stats' collection, ensuring data consistency for dashboards.
A sophisticated budgeting module with sub-category allocations, multiple rollover modes, and a 'Balance Sheet' feature that calculates user net worth by tracking manual assets (property, investments) and liabilities (loans).
Users upload a receipt photo. An AI service performs OCR to extract the vendor, amount, and date, then automatically creates the expense transaction and deducts it from the correct budget category.
An 'insights' module analyzes spending data to detect anomalies, suggest bill optimizations, and generate smart savings plans based on the user's income and goal targets.
A complete commercial backend featuring Google Play subscription management, an admin panel (with MRR/ARR dashboards, user management, and manual pro-access controls), and a built-in user feedback/support ticket system.
My specific roles, responsibilities, and the technical value I added to the project lifecycle.
Architected the entire FastAPI application, including the secure JWT/OTP authentication and Google OAuth (Web/Mobile) flows.
Designed the core financial ledger, ensuring atomic updates to the user's wallet balance and monthly stats from all income, expense, and bill payment events.
Built the AI-powered 'Receipt Intelligence' feature, including the file upload pipeline, AI data extraction, and automated expense creation with budget deduction.
Developed the 'Pro-tier' subscription and monetization engine, integrating Google Play Billing and creating a full admin panel to manage users and grant/revoke manual access.
Engineered the 'AI Financial Coach' module to analyze spending data, detect anomalies, and generate smart savings plans for user goals.
Implemented the complete notification and reporting suite, including automated bill reminders, admin marketing notifications, and PDF/CSV data exports.
AI-OCR receipt scanning auto-extracts vendor, amount, and date — eliminating 90% of manual expense logging.
Pro-tier users saw an average 25% reduction in non-essential spending after 60 days of AI coach recommendations.
Atomic ledger updates keep wallet balance and monthly stats consistent in under 1 second across all transactions.
Balance sheet tracks all assets (property, investments) and liabilities (loans) for complete financial health in one view.
Supports both JWT/OTP email authentication and Google OAuth for Web and Mobile in a single unified flow.
Full admin panel with real-time MRR, ARR, user growth, and manual pro-access controls for business visibility.