Huzifa Noor

Collaborated remotely with an Agri-fintech startup based in South Africa from the United States to build machine learning
model in AzureML which improved weather forecast, reduced water usage, and supported smarter farming decision in the
African regions.
• Designed and deployed AzureML pipelines connected to Blob storage to pull in live weather, which improved forecast
accuracy by 35%.
• Refactored over 200 lines of Python scripts to automate the streamlining of weather data ingestion and processing,
resulting in 76% boost in efficiency.
• Developed a reinforcement learning model in AzureML studio to help optimize irrigation schedules, reducing water
usage by one third without compromising crop yield.

Assisted Northwestern University’s MCDC team on a project for FracTracker Alliance by scraping and cleaning fracking data,
including images, and using ArcGIS Pro to create maps that made it easy for stakeholders to spot trends and differences across
regions.
• Perform data analysis for FracTracker Alliance, uncovering critical insights from national oil and gas fracking data that
shaped environmentally informed policy decision.
• Processes and cleaned large datasets in Jupyter Notebooks, boosting machine learning accuracy by 88% through pre-
processing techniques.
• Created interactive GIS-based visualization to map regional fracking activity, enabling stakeholders to easily spot
trends, detecting difference, and make data-driven decisions.