SheWealth AI

• 

  A completed dashboard

Inspiration

Women own only 20% of global wealth — not because they lack ambition, but because no one ever built a financial tool that truly understood them.

Generic finance apps ignore the realities women face: career breaks, gender pay gaps, irregular freelance income, and the pressure to prioritize others over personal savings. A student saving ₹500/month, a homemaker managing household budgets, and a working professional building a startup fund all have completely different needs — yet most apps treat them the same.

"What if there was a financial companion that actually got her?"

That question became SheWealth AI — a smart, judgment-free financial coach built exclusively for women at every life stage.

What it does

SheWealth AI is a full-stack MERN + Gemini AI web app that delivers genuinely personalized financial guidance:

How I built it

SheWealth AI is a decoupled full-stack architecture:

• Frontend — React.js + Tailwind CSS for the dashboard, charts, simulator, and SheBot chat UI
• Backend — Node.js + Express REST API handling auth, goals, and transaction routes
• Database — MongoDB + Mongoose for user profiles, transactions, and savings history
• AI Engine — Google Gemini 1.5 Flash with custom system prompts for domain-specific financial advice
• Auth — JWT-based secure login with life stage onboarding flow
• Charts — Recharts for income/expense bars and SIP growth projection
• Deployment — Render for live backend + frontend hosting SIP Formula Used

The Investment Simulator uses the standard SIP compound interest formula:

$$M = P \times \left[\frac{(1 + r)^n - 1}{r}\right] \times (1 + r)$$

Where P = monthly investment, r = monthly interest rate, n = total months

Challenges we ran into

• AI personalization across life stages — crafting Gemini system prompts that feel genuinely relevant whether you're a student or a homemaker, without being generic
• Securing API keys in a MERN deployment — moved all credentials to server-side environment variables on Render, never exposed in the React frontend
• Making finance approachable — iterated on the dark luxury UI to feel empowering, not intimidating, for first-time investors
• SIP simulator accuracy — validated compound interest outputs against real-world SIP calculators to ensure trustworthy projections

Accomplishments that we're proud of

✅ Built a fully functional AI financial dashboard within the hackathon timeline

• ✅ Real SIP simulator with live sliders and dynamic growth projection charts
• ✅ Gemini AI chatbot that gives genuinely useful, context-aware advice
• ✅ Inclusive onboarding covering 4 distinct women's life stages
• ✅ Production-ready dark UI that makes finance feel elegant & accessible

What i learned

• Architecting a clean MERN stack app with proper separation of concerns
• Integrating Google Gemini AI with custom domain-specific system prompts
• Designing for real user personas — a student and a homemaker need very different financial features
• Secure environment-based API key management for cloud deployments
• Making data visualizations approachable for non-technical users

What's next for SheWealth AI

• Architecting a clean MERN stack app with proper separation of concerns
• Integrating Google Gemini AI with custom domain-specific system prompts
• Designing for real user personas — a student and a homemaker need very different financial features
• Secure environment-based API key management for cloud deployments
• Making data visualizations approachable for non-technical users

Built With

• api
• css
• express.js
• gemini
• google
• javascript
• jwt
• mongodb
• node.js
• react.js
• recharts
• render
• tailwind