Experience

Controls Software and Simulation Engineer Ithaca, NY, US

Worked as part of the team for Project Orion - a NASA-funded project for building UTM for advanced air mobility - This project can be found at projectorion.info
Engineered a PC5 broadcast-style radio communication stack enabling low-latency, drone-to-drone coordination in airspace, including UDP/PC5 proxies and real-time telemetry synchronization.
Developed an aerial navigation and collision-avoidance system using Unity simulation, PX4 flight stack, and ROS 2 to avoid obstacles, predict trajectories of nearby drones, and perform safe multi-agent maneuvers.
Trained navigation transformers to optimize training performance for larger-scale applications as part of an exclusive NVIDIA Hackathon, resulting in improved UAV coordination.

Designed and implemented a high-performance MPC locomotion framework for a quadruped robot using Drake, enabling stable autonomous navigation with explicit contact modeling over varieties of terrain.
Built a training environment in Python for multi-agent reinforcement learning agents, enabling tasks completion under a decentralized coordination framework.
Ran experiments comparing multiple simulation platforms (Unity, PyGame, Lua) to optimize multi-agent learning performance and platform fidelity.

Led a 70-member interdisciplinary team (mechanical, electrical, software) competing in the University Rover Challenge, overseeing subsystem integration and guiding development of electro-mechanical rover controls in ROS2.
Developed inverse kinematics algorithms for the robotic arm, enabling reliable teleoperation via a PS5 controller for dexterous tasks.
Architected RoverNet, a hardware-software integration driver responsible for deciphering top-level software commands into transmittable CAN-FD frames.
Developed ROS 2 sensor integration pipelines, interfacing with GPS, LiDAR, IMU, and camera APIs to expose synchronized, autonomy-ready data streams used for terrain mapping and obstacle avoidance.