Sushma S

I am a meticulous data scientist with academic training in statistics, machine learning, and signal processing, combined with hands-on experience in computer vision, time-series forecasting, and reinforcement learning. I have implemented transformer-based classifiers, synthetic image augmentation using Blender, LSTM forecasting for energy management, and real-time vision systems, demonstrating strong problem-solving and reproducible project practices.

I have contributed to research projects at UMass Amherst and completed an internship augmenting COCO datasets and building few-shot image classifiers with high accuracy. I seek collaborative teams where I can apply my analytical rigor, software skills (Python, TensorFlow, PyTorch, SQL, AWS) and domain knowledge to deliver production-ready models and measurable impact.