Robotics and Machine Learning Engineer

Engineered and optimized Sim2Real and Real2Sim pipelines using Isaac Sim and Isaac Lab for complex robotic manipulation tasks. Set up robot demonstration collection and data pipelines supporting model training and evaluation.

Fine-tuned, evaluated, and benchmarked VLA models including π0.5, SmolVLA, Groot N1.5, and ACT across parallel simulation environments and physical robotic platforms. Built automated evaluation pipelines with LeRobot and PyTorch to assess policy behavior, zero-shot generalization, trajectory quality, and robustness under changing conditions.

Developed and deployed real-time deep learning perception systems for pickleball tracking, including custom object detection and object tracking pipelines. Built an eye-gaze-based cursor control system for assistive human-computer interaction.

Integrated vision-based machine learning models into ROS-based systems and IoT edge pipelines for perception-driven closed-loop robotic control. Programmed rigid-body dynamic simulations for multi-joint manipulation tasks using PyBullet and OpenAI Gym to benchmark reinforcement learning agents.

Developed and trained convolutional neural networks for ear-health medical classification using otoscopic image data.

Optimized and integrated deep learning inference pipelines into Android-based systems to support point-of-care diagnostics.

Contributed to autonomous localization stacks by implementing SLAM-based state estimation using fiducial markers and TagSLAM.

Engineered and configured ROS node architectures for real-time sensor-stream mapping, odometry tracking, and 6-DoF pose estimation.

Completed advanced coursework and projects in artificial intelligence, robotics, machine learning, computer vision, LLMs, VLMs, and multimodal foundation models.

Specialized in Automation and Robotics. Graduated with a CGPA of 8.02/10.