InvestEd is an AI-powered financial literacy app that transforms confusing bank statements into fun, engaging, and personalized explainer videos.
Built at Hack the North 2025, our goal was to help students (and beyond) understand their spending habits, save smarter, and invest toward their "my first..." goals — laptop, car, trip, and more.
As university students, we know how overwhelming personal finance can feel. Charts, statements, and jargon rarely help — they only add stress.
We wanted to make saving and investing accessible, engaging, and educational. By combining AI + video storytelling, InvestEd gives your money a voice.
🧾 76% of Canadians struggle with saving consistently, and Millennials/Gen Z overspend by ~15% monthly vs. their budgets.
InvestEd helps close that gap.
- AI Finance Assistant (Portfolius): Explains your spending habits in plain English.
- Personalized Explainer Videos: Automatically generated with Manim, breaking down overspending and showing how small changes build big savings.
- Interactive Web App: Clean Next.js frontend where users can view insights, generate videos, and replay them anytime.
- Positive Nudges: Instead of guilt, InvestEd reframes overspending as opportunities to save & invest.
- Frontend: Next.js + TailwindCSS → Clean, banking-inspired UI.
- Backend: FastAPI (Python) → Orchestrates video generation pipeline.
- Cohere LLM: Generates short, simple scripts explaining overspending & investment tips.
- Manim Renderer: Converts scripts into animated explainer videos in under 30 seconds.
- Supabase Storage: Stores and serves large MP4 files securely via public/signed URLs.
- Deployment: Frontend on Vercel; Backend on Render/Heroku.
- CORS issues & API rewrites during frontend ↔ backend integration.
- Optimizing Manim video rendering (initially minutes → now sub-30s).
- Handling large MP4 storage & retrieval with Supabase.
- Coordinating multiple independent components into one seamless pipeline.
- Built a full end-to-end pipeline: Cohere → Manim → Supabase → Frontend in <30s.
- Designed a seamless, banking-style interface inspired by real finance apps.
- Learned to optimize AI prompts, rendering configs, and storage integration for hackathon constraints.
- Delivered a working demo that turns raw spending data into a clear, animated video.
- Prompt engineering is critical for syncing AI-generated scripts with video timing.
- Pipeline orchestration (status polling, error handling) keeps UX smooth.
- Rendering optimizations (fps, resolution, animation simplification) can make or break performance.
- Team integration matters: constant communication = faster debugging.
- Gamified learning: Streaks & rewards for saving more consistently.
- Cloud scaling: Offload heavy rendering to worker nodes for parallel video generation.
- Mobile apps: Bring InvestEd to iOS & Android for anywhere access.
 |
 |
 |
 |
- Node.js 18+
- Python 3.10+
- Supabase account
- Cohere API key
# Clone repo
git clone https://github.com/<your-org>/invested.git
cd invested
# Frontend setup
cd frontend
npm install
Create .env files in frontend/ and backend/ with:
npm run dev
# Backend setup
cd ../backend
pip install -r requirements.txt
uvicorn app:app --reload --port 8000
# Frontend
NEXT_PUBLIC_BACKEND_URL=http://127.0.0.1:8000
# Backend
COHERE_API_KEY=your_cohere_api_key
SUPABASE_URL=your_supabase_url
SUPABASE_KEY=your_supabase_key
-
Rishi — Backend (Cohere integration, API orchestration)
-
Hreem — Frontend design & implementation
-
Hetarth — Video storage & analytics integration
-
Ayman — Manim optimization & rendering pipeline
Cohere for LLM API access.
Manim for rendering videos.
Supabase for storage & database support.
Hack the North 2025 organizers & mentors.