Aquatic Sustainability

This started as a GT Big Data project around a real gap: in many less-developed regions there's too little infrastructure and data to catch contamination, predict floods or droughts, or track whether water stays affordable. The idea was to aggregate satellite imagery and historical datasets into one interface that geolocates watershed anomalies and inequities, so citizens and governments can act earlier. (Team effort — my specific contribution is flagged separately in the role field.)

Stack

• Python
• Jupyter
• Computer vision (CNNs)
• Time-series (RNN/LSTM)