Tayyibe Gunduz is an AI/ML Engineer with a strong background in computer science. Her work focuses on the development of neural network compression algorithms and their practical implementation in machine learning systems. She also contributes to the design and development of scalable federated learning frameworks aimed at enabling efficient and distributed model training.
At Pangea, Tayyibe contributes to an NSF-funded project focused on developing an AI reliability platform designed to assess and improve the robustness of machine learning systems. Her work includes analyzing and developing compressed yet reliable neural network models that maintain strong predictive performance while reducing computational complexity. She also contributes to research exploring the use of hyperbolic geometry and other advanced mathematical frameworks to develop more efficient representations for high-dimensional data.
In addition to her work on AI reliability systems, Tayyibe has experience developing machine learning models for agricultural and environmental applications, including deep learning models for vegetation recognition using drone-based aerial imagery. She has contributed to improving model performance through enhanced image preprocessing and data augmentation, as well as integrating model outputs into geographic information systems (GIS) to support spatial analysis and decision-making in agricultural monitoring and land management.