Track

best overal, LAKOG, Memenome Best Don Pollo-themed hack, i’m just a chill guy (energy track), most centered div

What it does

SolarCalc is an application designed to help users calculate the potential energy savings from installing solar panels. It takes inputs like monthly energy consumption, system size, installation costs, sunlight hours, battery costs, and financing options, and provides an estimate of energy savings, payback period, and the potential benefits of adding battery storage.

How we built it

We built SolarCalc using Flask, a lightweight web framework for Python, combined with the Hugging Face transformers library for implementing a chatbot feature. The app uses JavaScript to handle the frontend interactions, sending user inputs to the Flask backend, where calculations are performed. Additionally, we use HTML and CSS for the user interface. The core calculations involve estimating energy savings and payback periods based on various inputs.

Challenges we ran into

The initial integration of the Hugging Face model for conversational AI posed some challenges in ensuring it worked seamlessly with the frontend. We needed to ensure that the solar savings, payback period, and other financial estimates were both accurate and understandable for users.

Accomplishments that we're proud of

We added a conversational chatbot feature using Hugging Face's models, allowing users to interact with the app and get answers to questions related to solar energy.

What we learned

Flask Web Development: We gained hands-on experience with Flask for building a web-based application and integrating backend logic with a frontend user interface. Machine Learning Model Integration: We learned how to integrate machine learning models (like Hugging Face’s GPT models) into a web application and use them for real-time interactions with users. Handling Complex Calculations: We learned how to handle complex calculations, particularly related to energy savings and solar installation costs, ensuring they are accurate and user-friendly.

Built With

Share this project:

Updates