Recent projects
Urban Sustainability Score — Drivers, Trade-offs, and Scenario Simulator
An advanced applied analytics project that models how urban form, infrastructure, and environmental factors interact to shape a composite urban sustainability score. Built with Python, Pandas, scikit-learn, Plotly, and Streamlit, the project combines predictive modeling, model interpretability, and interactive scenario simulation to explore policy-relevant trade-offs across cities. The workflow includes feature engineering across built-environment and environmental indicators, supervised learning for sustainability scoring, sensitivity analysis, and counterfactual “what-if” scenarios. A fully interactive Streamlit dashboard allows users to adjust key drivers (e.g., green cover, transit access, renewable energy use) and immediately observe impacts on predicted sustainability outcomes, supporting exploratory analysis and decision-oriented insight.
Canada Climate (1940–2020): Long-Run Trends & Extremes
An end-to-end climate analytics project examining 80 years of city-level climate data across Canada, with a focus on long-run temperature trends, seasonal shifts, and climate extremes. Built with Python, Pandas, Plotly, and Streamlit, the project integrates robust data processing with advanced interactive visualizations. The workflow includes daily-to-monthly and yearly rollups, baseline climate normal comparisons (1961–1990 vs 1991–2020), trend estimation, distributional analysis, and spatial mapping of warming signals and extremes across major Canadian cities. Results are delivered through an interactive Streamlit dashboard designed for exploratory analysis and communication of climate patterns.
Toronto AOC Water-Quality Analytics
An end-to-end analysis of Toronto’s AOC1 water-quality monitoring network. Built with Python, Pandas, Seaborn, and Streamlit, this project explores operational water datasets to uncover trends, anomalies, and spatio-temporal patterns.The workflow includes data cleaning, feature extraction, exploratory analysis, temporal trends, anomaly detection, and spatial clustering across monitoring stations. The results are presented through a fully interactive Streamlit dashboard. Key highlights:Multi-stage data processing (cleaning, EDA, modelling, export). Trend analysis for high-priority parameters. Anomaly and outlier detection in time-series signals. Spatial hotspot identification for water-quality indicators.Interactive dashboard for exploration and reporting.
Ontario Permit-to-Take-Water (PTTW) Analytics Project
I built a full end-to-end analytical workflow to explore Ontario’s Permit-to-Take-Water dataset, including data cleaning, feature engineering, temporal trends, sector profiles, and spatial mapping. The final output is an interactive Streamlit dashboard that visualizes water-taking activity across the province and highlights key insights such as high-volume permits, sector patterns, and geographic hotspots.
California Air Quality Dashboard
An interactive dashboard visualizing air quality trends across California, including NOâ‚‚ and CO concentrations. Built with Python, Plotly, and Streamlit to highlight temporal patterns and city-level insights.
This dashboard analyzes California’s air quality using open environmental data, focusing on pollutants such as NOâ‚‚ and CO. It provides dynamic time-series exploration, sensor-level trends, and clean visual storytelling through Plotly and Streamlit. Designed to help users quickly identify patterns, anomalies, and long-term shifts in air quality.
​
​
NOAA NYC Climate Analysis (2020–2025)
This project analyzes five years of NOAA GSOM climate data for the New York City metro area, focusing on temperature, precipitation, and seasonal trends. The goal was to explore how local climate indicators have shifted over recent years and to highlight meaningful patterns using interactive charts and time-series visuals. The dashboard presents trends, anomalies, and comparisons in a clear and intuitive layout built with Python, Pandas, Plotly, and Streamlit. It gives users an accessible way to examine warming patterns, extreme weather shifts, and long-term climate behavior in NYC.
Canada GHG Emissions Projections Dashboard
This project is an interactive dashboard that visualizes Canada’s greenhouse-gas emissions projections by scenario, sector, and province using publicly available federal data. Built with Python, Plotly, and Streamlit, it allows users to explore long-term emissions pathways, compare sector trends, and examine how different policy scenarios influence national and regional outcomes. The dashboard focuses on clear data storytelling through animated charts, interactive filters, and comparisons that help users understand Canada’s emissions landscape and potential future trajectories.
Understanding U.S. Urban Areas: Spatial Scale & Urban Land Concentration
An analytical exploration of U.S. Census Urbanized Areas (UAs) and Urban Clusters (UCs) focused on urban scale, spatial concentration, and long-tail dominance. This project examines how urban land is distributed across thousands of urban areas, revealing strong inequality where a small number of very large metropolitan regions account for a disproportionate share of total urban land. Built with Python, Pandas, Plotly, and Streamlit, the analysis combines advanced distributional analytics (log-scaled boxplots, long-tail and outlier analysis) with GIS-informed spatial visualization. The interactive dashboard enables users to explore size-based typologies, compare UAs versus UCs, identify extreme-scale urban areas, and interpret spatial patterns through density-based mapping.
