I’m an AI researcher working across theoretical and empirical machine learning. I develop principles and algorithms for AI systems that learn more from limited data, generate with less computation, and adapt to the people they assist. My research connects learning and optimization with generative modeling and human–AI collaboration, toward a broader goal of making AI more practicable: not simply more capable, but more effective under the constraints that matter in practice.

My work combines mathematical analysis, algorithm design, and experiments with models and human–AI teams. Recent projects have developed principles for data-efficient language-model training, methods for high-quality diffusion language generation in fewer steps, and adaptive AI ensembles that account for human expertise and reliance.

I earned my PhD and MS in Computer Science at Purdue University, where I was fortunate to be advised by Ming Yin and Rajiv Khanna. Before Purdue, I studied Electrical Engineering, with a minor in Computer Science, at LUMS in Lahore, Pakistan.