Mitalee Garg

Organized and executed more than 15+ social and cultural events, fostering an inclusive environment for international students. Directed and coordinated 4 diverse teams, enhancing collaboration, which reflects enthusiasm and creative approaches.

Project Lead for Defect detection deep learning pipeline on distributed systems, achieving 96% accuracy by implementing models such as YOLO and increasing 47% efficiency. Conducted comprehensive research with on scalable pre-training and post-training pipelines, data preparation, and research on various GenAI models, leading to 15% faster processing.

Streamlined IT communications across global teams and enhanced internal tooling, demonstrating problem-solving and process optimization in a large organization.. Enhanced inter-departmental websites through efficient HTML5 updates. Leveraged SAP and Docker to support data analysis, contributing to efficient information flow and technical reporting.

Thesis "A Differentiable Threshold Optimization Framework for Transformer Models with Application to Geospatial Data" Developed AI models for geospatial data analysis applicable to logistic monitoring and optimization of parcel delivery for e-commerce.

Thesis "Automated Driving Systems with Deep Learning and deployed the model on Nvidia GPU"

Thesis "Automated Driving Systems with Deep Learning and deployed the model on Nvidia GPU"