Inspiration
Rising energy costs and climate concerns inspired us to create WattWise—a platform that empowers households and businesses to not only understand but also act on their energy usage. We were motivated by the vision of sustainable living and the role AI can play in shaping smarter, greener cities.
What it does
WattWise is an AI-powered energy management platform that:
- Tracks real-time energy consumption, costs, and CO₂ emissions.
- Forecasts usage for the next 24 hours using XGBoost, LSTM, and Prophet.
- Provides personalized recommendations and alerts for efficiency.
- Simulates realistic data streams to demonstrate energy-saving potential.
- Engages users with gamified goals and achievements.
How we built it
- Backend: Flask with Flask-SQLAlchemy and Flask-Login.
- Frontend: HTML, CSS, JavaScript, Bootstrap, and Chart.js for visualization.
- Database: MySQL for production and SQLite for testing.
- ML Pipeline: Data preprocessing with Pandas/NumPy, forecasting with XGBoost, LSTM, and Prophet.
- Integrated the models into Flask via dedicated utilities and an
/api/forecastendpoint.
Challenges we ran into
- Integrating ML models into Flask endpoints without blocking performance.
- Handling inconsistencies in the UCI dataset and reshaping input data for LSTM.
- Debugging SQLAlchemy migrations and ensuring smooth database initialization.
Accomplishments that we're proud of
- Built a full-stack, AI-driven energy forecasting app in a hackathon timeframe.
- Successfully deployed multiple ML models and compared them in production.
- Designed a user-friendly dashboard with real-time interactivity.
What we learned
- Best practices for merging machine learning with production-grade web apps.
- Trade-offs between time-series forecasting models (e.g., Prophet vs LSTM).
- Practical database management and API integration under time constraints.
- How gamification and UX design can boost user adoption of sustainability tools.
What's next for WattWise
- Expand device management to IoT hardware integration.
- Incorporate reinforcement learning for adaptive energy recommendations.
- Add payment/credit systems to reward users for reducing consumption.
- Deploy WattWise on cloud platforms for broader scalability.
Log in or sign up for Devpost to join the conversation.