## 💡 Inspiration

Every year, millions of students dream of studying abroad — but the process is 
overwhelming, expensive to navigate, and painfully opaque. We watched friends 
spend months researching universities, guessing their admission chances, and 
making six-figure financial decisions with almost no data to back them up.

Existing platforms either show you a list of universities or charge consultants 
₹50,000+ for guidance that should be accessible to everyone. We asked: *what if 
an AI could do all of that — instantly, for free, and make it actually fun?*

That's EduPilot.

---

## 🏗️ How We Built It

EduPilot is a full-stack AI platform built on:

- **Frontend:** React 18 + Vite + Tailwind CSS + Framer Motion for a 
  premium glassmorphism UI
- **Backend:** FastAPI (Python) + Uvicorn serving AI logic and data APIs
- **AI Engine:** Google Gemini API powering admission probability analysis, 
  essay coaching, and smart university recommendations
- **Database:** Supabase (PostgreSQL) for user profiles, progress tracking, 
  and real-time leaderboard sync

The gamification engine tracks user progress across missions — research, 
application, financial planning — and maps them to a leveling system from 
*Elite Navigator* to *Global Scholar*.

---

## 🚧 Challenges We Faced

**Admission probability is not a solved problem.** Universities don't publish 
their exact selection criteria, so we had to design a model that combines 
user profile data with Gemini's grounded knowledge of historical admit patterns 
— and communicate uncertainty honestly rather than giving false confidence.

**Real-time gamification at scale** required careful database architecture. 
Keeping leaderboard scores, quest completions, and level progression in sync 
across sessions without latency meant restructuring our Supabase schema multiple 
times.

**UI performance** with heavy Framer Motion animations on data-dense dashboards 
caused frame drops on mid-range devices — solved by lazy-loading panels and 
deferring non-critical animations.

---

## 📚 What We Learned

- How to design AI outputs that feel trustworthy rather than authoritative — 
  especially for high-stakes decisions like university applications
- That gamification works best when the rewards map directly to real progress, 
  not arbitrary points
- FastAPI + Supabase is a genuinely fast stack for AI-backed applications with 
  real-time requirements

Built With

  • fastapi
  • framer-motion
  • google-gemini-api
  • lucide-react
  • postgresql
  • python
  • react-18
  • supabase
  • tailwind-css
  • uvicorn
  • vite
Share this project:

Updates