Senior Roboticist at RoboStore | Robotics Engineer & Educator — Building intelligent humanoid systems with the Unitree G1, developing custom RL policies, and creating multimodal AI frameworks for embodied intelligence. Course creator at RoboUniversity.

• Senior Roboticist at RoboStore — developing advanced humanoid robotics systems, web-based robotics tools, and simulation platforms
• Building custom policy training pipelines for the Unitree G1 humanoid — locomotion, manipulation, and whole-body control via reinforcement learning
• Developing custom multimodal LLMs for embodied AI — integrating vision, language, and proprioceptive sensing on humanoid platforms
• Created Robot IDE — a browser-based platform for visualizing, animating, and programming robots with 3D model loading (URDF, MuJoCo XML, USD, STL, OBJ, DAE), real-time MuJoCo physics simulation, AI policy execution (ONNX), motion library, motion planner, built-in code editor, and sequence builder
• Built a web-based motion & physics simulator for the Unitree G1 — interactive motion planning, physics visualization, and joint manipulation in the browser
• Creating the RoboUniversity course catalog — full-stack humanoid robotics education from setup to Sim2Real deployment
• Working with OM1 Agent Architecture, ROS2, Isaac Sim, MuJoCo, and NVIDIA Jetson Orin for real-world humanoid intelligence
• Published 100+ iOS apps (135K+ downloads) through Feraco Media — including AI/ML-powered tools built with ResearchKit, ARKit, and CoreML
• Co-authored the DroneBlocks Robotics 101: Land. Air. Sea. curriculum — a year-long project-based STEM program across ground, aerial, and underwater platforms
• Collaborated with Stanford University to develop BActivate, a digital health app integrating VR for behavioral activation therapy
• Founded a Coral Aquaculture Research Program and built CoreML-powered marine species identification (350+ species)
• Science Research Teacher & Cybersecurity Instructor at Walt Whitman High School (2015-present) | Formerly Adjunct Professor of Drones at Five Towns College
• Open to collaborations in Humanoid Robotics, Embodied AI, RL Policy Training, Multimodal LLMs, Sim2Real Transfer

RL Policy Training for G1

Multimodal AI Agent Platform

Humanoid Motion Data

Multimodal LLM for G1

Developing end-to-end reinforcement learning pipelines for the Unitree G1 humanoid:

• Locomotion policies — PPO/SAC-based training for stable bipedal walking, running, and stair traversal
• Whole-body control — Joint-level policy training with MuJoCo simulation and Isaac Sim environments
• Sim2Real transfer — Bridging simulation-trained policies to physical G1 hardware via domain randomization and system identification
• Custom reward shaping — Task-specific reward functions for manipulation, navigation, and human-robot interaction

Building custom large language models that integrate multiple modalities for humanoid intelligence:

• Vision-Language-Action (VLA) models — Connecting camera perception to high-level reasoning and low-level motor control on the G1
• Multimodal sensor fusion — Integrating RGB/depth vision, IMU, force/torque, and proprioceptive data into unified LLM representations
• RAG-augmented robot memory — Retrieval-augmented generation for contextual, long-horizon task planning
• Real-time inference on edge — Optimized model deployment on NVIDIA Jetson Orin for on-robot multimodal reasoning

Building the full-stack humanoid robotics education platform at RoboUniversity

Skills taught across the platform: OM1 Agent Architecture · ROS2 & CycloneDDS · Low-Level G1 SDK Motor Control · PPO Reinforcement Learning · Isaac Sim / Isaac Lab · MuJoCo Simulation · Teleoperation & VR Streaming · Python / uv Tooling · Sim2Real Transfer · Multimodal Sensor Fusion · Domain Randomization · NVIDIA Jetson Orin Deployment

Python	C+	Swift	JavaScript	TypeScript	Rust	Bash
PyTorch	TensorFlow	OpenCV	ROS2	Arduino	Raspberry Pi	MuJoCo
Docker	AWS	Linux	Isaac Sim	Jetson Orin	GitHub	VS Code