I am an Applied Scientist with a Ph.D. in Transportation Systems from New York University, specializing in large-scale optimization, machine learning, and LLM-based systems.
I build production-oriented ML and optimization solutions, including reinforcement learning methods for vehicle routing, prompt optimization frameworks, and automated root-cause diagnostics for large-scale agent failures.
My technical background includes mathematical optimization, mixed-integer programming, deep learning, spatiotemporal modeling, and routing heuristics. I have worked with GCNs, self-attention, GRUs, LSTMs, autoencoders, Bayesian optimization, and commercial vehicle-routing solvers.
At Amazon, I apply reinforcement learning to improve column-generation workflows for vehicle routing and develop LLM-based diagnostic systems using CloudWatch and DynamoDB data. My work has achieved more than 99% root-cause analysis accuracy.
I also bring research and industry experience in transportation systems, mobility analytics, facility location, energy planning, and food-delivery optimization. My work has contributed to peer-reviewed publications and measurable operational savings.
I am experienced with Python, PyTorch, Gurobi, Pyomo, PySpark, scikit-learn, AWS SageMaker, and AWS Bedrock, and I enjoy translating advanced research into scalable practical systems.
