Nistha Gautam

• ● Improved 5G tower placement decisions for disaster response using ETL pipelines in Python, unifying four global data sources on geospatial, altitude, signal, and disaster data, warehoused in AWS S3.

• ● Predicted 5G signal coverage in disaster-prone regions using engineered features and scikit-learn model for continuous drone connectivity during rescues.

• ● Delivered clear interactive Plotly visualizations, helping stakeholders quickly identify optimal locations for new 5G tower deployment.

• ● Led a four member team to analyze 3GB high-resolution breast tissue slides, building a PyTorch/OpenCV model to predict tumour regions while handling massive image data efficiently.

• ● Developed a Django web interface with PostgreSQL for backend, enabling slide input, patch visualization, and immunity scoring with reduced computation and improved latency.

• ● Managed project lifecycle, liaising with client and supervisors, resolving team conflicts, delegating tasks, and delivering predictive analytics MVP.

• ● Experimented with a bare-bones diffusion model, modifying architecture and training steps to recreate a DALL·E-style image generation pipeline and understand generative model internals.

• ● Researched distributed random forest IDS, using synthetic DDoS datasets to detect anomalies and improve accuracy and scalability in large-scale networks.