About
I am an AI Research Scientist at American Express, where I develop machine learning methods for learning spatiotemporal and relational representations from complex, high-volume transactional data.
My work focuses on building foundational models for prediction and compression. Before joining American Express, I held research and engineering roles at Imperial College London, Stanford University, Huawei Technologies R&D, Vodafone, Bilkent University, Databoss Security & Analytics, and Leo Augmented Reality. Across these roles, I worked on compression, natural language processing, multimodal learning, computer vision and anomaly detection.
I also developed PySAD, an open-source Python framework for anomaly detection on streaming data, and have contributed to 80+ open-source projects. My research, including 8 journal papers and 12 conference papers, has been published in leading IEEE journals and conferences.
I am particularly interested in translating advanced machine learning research into practical systems that are scalable, robust, privacy-aware, and impactful in data-intensive domains such as finance, technology, and intelligent infrastructure.