Backend team project
Dublin Bikes
A web app for exploring bike station availability, weather, historical trends, and predicted occupancy.
Python · Flask · MySQL · scikit-learn
Overview
This team project combines a Flask application, MySQL data storage, external station and weather APIs, and an interactive map. Visitors can check station availability and explore historical occupancy alongside current weather information.
My backend work
I spent most of my time on the backend. The project collects station data from JCDecaux and weather data from OpenWeather, stores records in MySQL, and serves the data through Flask routes for the web interface.
Prediction work
I also contributed to a smaller part of the prediction work. The application uses a scikit-learn model trained on historical bike occupancy and weather-related features to estimate availability. The Flask backend exposes those predictions to the interface.