CityPulse

CityPulse, built by Muntasir and I, is an interactive dashboard for visualizing and predicting Toronto 311 service requests. Inspired by our shared passion for civic data, we designed a Flask backend to serve processed data and machine learning predictions and a modern frontend using Plotly.js for real-time visualizations. We automated data ingestion with a Python data pipeline that cleans raw CSVs, trains a predictive model (model.joblib), and exposes insights via a REST API. Along the way, we optimized our ML model to accurately forecast request completions, integrated responsive charts by ward, status, and time of day, and overcame creative challenges for the dashboard. Full course loads with a two-person team taught us to collaborate tightly on performance tuning and testing, resulting in a dashboard that empowers city planners and residents with actionable KPIs and trend analyses.

Built With

• css
• html
• javascript
• machine-learning
• matplotlib
• numpy
• pandas
• python
• scikit-learn