Develops LLM and multimodal inference frameworks, including SGLang, vLLM, and ATOM, for agentic AI and distributed serving.

Developed computer vision and imaging algorithms for image signal processors, Intel RealSense depth cameras, robotics, and automotive applications.

Researched deep-learning-based surrogate modeling, simulation, and image processing in the DB Lab. Also served as a tutorial lecturer for an Artificial Intelligence course.

Researched semantic segmentation and frameworks for accelerating deep learning workloads.

Conducted theoretical research in reinforcement learning, stochastic geometry, and convex optimization.