MetuchenFinAI — Municipal Financial Intelligence for the Borough of Metuchen, NJ
💡 Inspiration
Small municipalities operate with limited financial staff and outdated spreadsheets, yet they manage millions in taxpayer dollars across investments, operating budgets, capital projects, and public services. We wanted to give a mayor the same caliber of financial intelligence that a Fortune 500 CFO takes for granted — built specifically for local government.
Metuchen, NJ is a borough of ~15,300 residents with a ~$10.75M investment portfolio and a century of spending history. We saw an opportunity to turn that raw municipal data into an interactive, AI-augmented decision console.
📊 What It Does
MetuchenFinAI is a financial intelligence dashboard that unifies the borough's entire fiscal picture into one console:
- Portfolio Overview — Live investment holdings across cash, fixed income, equities, and real estate, with weighted yield, annual income, and asset-class allocation visualizations
- Historical Expenses (1926–2026) — A century of spending data modeled across real economic eras: the Great Depression, postwar boom, 1970s inflation, 2008 recession, and COVID-19
- Revenue Analysis — Property tax (58%), state aid, federal grants, and growth trends
- Capital Projects — Budget vs. spent tracking for infrastructure investments
- Department Scorecards — Efficiency, on-time delivery, and budget utilization by department
- Budget Variance — Real-time over/under budget detection
- Demographics, Public Safety & Public Works — Civic context for spending decisions
- AI Assistant — A conversational assistant embedded in the app that answers questions about the town's finances, surfaces savings opportunities, and provides actionable guidance
🛠️ How We Built It
- React + Vite for a fast, componentized frontend
- Tailwind CSS with a financial-grade slate/navy design system
- Recharts for interactive data visualizations (area, bar, pie, radar, line charts)
- Base44 platform for backend-as-a-service: authentication, database entities, and serverless integrations
- LLM integration powering the in-app AI assistant with municipal finance context
- A custom sidebar layout with 9 dedicated analytical pages and a floating AI chat widget available on every screen
- Row-level security protecting financial data per authenticated user
🧗 Challenges We Ran Into
- Modeling a century of representative municipal expense data (1926–2026) that reflects real economic inflection points — we hand-crafted growth/contraction curves for each era
- Getting auth routing right for published apps (a missing login route caused a 404 in production)
- Balancing data density with readability — municipal finance has many dimensions, and we had to choose the right chart for each
🏆 Accomplishments We're Proud Of
- A cohesive 9-page console that feels like a real product, not a demo
- An AI assistant that understands the borough's specific financial context
- Historical data that actually tells a story — you can see the Depression, the inflation era, and COVID in the spending curve
📚 What We Learned
- Municipal finance is more nuanced than it looks — departments, debt service, capital vs. operating, and revenue interdependencies all matter
- Good data visualization is about choosing what not to show
- AI assistants are most useful when given domain-specific context, not generic instructions
🚀 What's Next
- CSV/data export for council meetings
- Real-time integration with municipal accounting systems
- Predictive budget modeling using historical trends
Built With
- ai
- assistant
- base44
- css
- data
- fintech
- gemini
- govtech
- javascript
- llm
- node.js
- react
- recharts
- responsive
- tailwind
- visualization
- vite
Log in or sign up for Devpost to join the conversation.