I am a Computer Science undergraduate with a strong focus on AI, machine learning, robotics, and systems-level engineering. I enjoy building end-to-end systems — from low-level embedded control to high-level learning algorithms.
I have worked on reinforcement learning, multi-agent systems, robotics perception, and embedded AI, including research-oriented projects and competitive engineering challenges. I value understanding concepts from first principles and translating theory into working systems.
I am building an autonomous robotic arm capable of mimicking human movements and learning to perform tasks through interaction. The focus is on combining perception, control, and learning-based approaches to enable adaptive and general-purpose manipulation.
Express interest to collaborateI am developing an autonomous rover or vehicle platform capable of independent navigation using a full autonomous navigation stack. The system focuses on perception, localization, mapping, planning, and control, enabling the rover to operate in real-world environments.
The platform is designed to be modular and customizable, allowing it to be adapted to solve domain-specific problems such as exploration, delivery, inspection, or research-oriented navigation tasks.
Express interest to collaborateDeveloped an AlphaZero-inspired self-learning framework combining Model-Based Reinforcement Learning and Monte-Carlo Tree Search (MCTS). The system improves through iterative self-play and generalizes to previously unseen states.
Built a full autonomous navigation stack for warehouse environments using ROS 2 and Nav2. Implemented frontier-based exploration, LiDAR point-cloud clustering (DBSCAN), QR-based metadata extraction, and YOLO-based object recognition on embedded hardware.
Designed a reinforcement learning pipeline for autonomous rover navigation on synthetic planetary terrains. Integrated generative models and scientific constraints to enable transfer to Moon and Mars-like datasets.
Developed an RL-based defense framework to mitigate adversarial jailbreak attempts in language models. The system learns safe response policies while preserving benign task performance.
Implemented GPU-parallelized simulations of black hole photon dynamics using CUDA. Focused on kernel optimization, memory efficiency, and numerical stability for high-performance computing.
Built a hands-free human–computer interaction system combining computer vision and speech recognition to enable iris-based cursor control and voice-driven automation.
Performed astrophysical photometry and clustering analysis on Hubble telescope data to study stellar populations and spatial structure within the NGC 1300 spiral galaxy.
Designed multiple embedded and robotic systems including a high-speed PID-controlled line follower, ESP32 cloud-controlled music system, and ROS 2-based drone swarm simulations with real-time visualization.
Built a full-stack student management platform using FastAPI, MongoDB, and SQL backends with secure authentication, modular services, and scalable data handling.