I am currently working on creating ML models to perform tasks for Unitree G1. I am open to collaborate on ML & Robotics research.

I am an AI & Robotics Engineer with a strong foundation in Machine Learning, Deep Learning, and Generative AI, focused on building intelligent systems that reason, learn, and act in the physical world. My work bridges ML algorithms, NLP models, and robotics, enabling robots to understand high-level instructions, learn complex behaviors, and operate autonomously in real environments.

I actively design and train machine learning and deep learning models for perception, decision-making, and control, with hands-on experience in reinforcement learning, imitation learning, and policy optimization for robotic tasks. A major part of my work involves addressing Sim2Sim and Sim2Real gaps, ensuring models trained in simulation generalize reliably to real robots.

In parallel, I explore NLP and Generative AI techniques—working with Large Language Models (LLMs) and Vision-Language Models (VLMs) to enable natural language understanding, task planning, and human-robot interaction. I am particularly interested in using GenAI for robotics, such as language-conditioned policies, multimodal reasoning, and autonomous task execution driven by natural language commands.

I build and deploy robotic applications using ROS / ROS 2, integrate ML models into real-time robotic systems, and leverage NVIDIA’s robotics stack for simulation, training, and deployment. I also develop web-based visualization and control interfaces, enabling intuitive monitoring and interaction with robots.

My goal is to push Physical AI forward by combining ML, NLP, and Generative AI with real-world robotic systems—transforming research ideas into practical, deployable solutions.