Zeshan is an AI and Machine Learning Engineer working on MIRA, an AI powered mental health resource chatbot that helps connect people to trusted services and information. He works across the full stack, building and maintaining the backend, frontend, embedding pipelines, and large language model integrations, while also managing GPU-based infrastructure and supporting deployments for partner organizations.

He holds a Master’s degree in Mathematics, Computer Science from the University of Waterloo and has a strong research background spanning federated learning, distributed systems, recommender systems, and applied machine learning. His work includes HERL, a reinforcement learning approach for optimizing homomorphic encryption in federated learning, as well as a widely cited survey on recommendation systems with over 500 citations and tens of thousands of downloads. He has also contributed to research in scalable systems and real-world challenges in federated learning.

He is particularly interested in large language models, federated learning, distributed machine learning systems, and building practical AI tools that make information more accessible and useful in real world settings.