Robotics & Cloud AI Engineer

Vijay Wulfekuhle

GitHub • 470-***-****• LinkedIn

EDUCATION

Kennesaw State University Marietta, GA May 2026

M.S in Robotics

Georgia State University Atlanta, GA May 2025

B.S in Computer Science

Amazon Web Services

Cloud Practitioner, AI Practitioner

EXPERIENCE

GreyOrange Lithia Springs, GA 05/2026 - Current

Robotics Operations and Maintenance Engineer

● Executed production-scale validation testing for a 50+ heterogeneous multi-agent autonomous mobile robot fleet, analyzing multi-agent coordination across staged fulfillment task execution

● Monitored fleet orchestration dashboards to detect execution faults, task-state inconsistencies, and exception conditions requiring distributed technical escalation for vertical retrieval and horizontal transfer systems

● Diagnosed and resolved AMR hardware/software faults through component replacement (cameras, motors) and electrical/log root cause analysis KSU Biomedical Lab Marietta, GA GitHub 11/2025 – 05/2026 Graduate Research Assistant

● Developed CAN bus actuation firmware in Python for a lower-limb exoskeleton, integrating CubeMars AK10-9 (knee) and MAB Robotics frameless MD-80 (ankle) brushless actuators across 4 joints on Raspberry Pi

● Implemented position, velocity, current, and braking control modes via extended CAN ID packet construction over RS -485

● Built a modular wheelchair exoskeleton that patients can interact with through GUI interaction of assistive and resistive control modes KSU Electric Vehicle Team Marietta, GA GitHub 05/2025 – 11/2025 Software Lead

● Fine-tuned DeepLabV3’s semantic segmentation model for lane detection on autonomous go-kart footage

● Calibrated 3D coordinates from stereo intrinsic and extrinsic camera calibration via the pin-hole method

● Developed a non-linear model predictive control algorithm and path planner for lane-keeping of go-kart throttle and steering

● Implemented on a custom embedded ground vehicle through ROS2 and BeamNG for simulation GSU Rocket Technologies Atlanta, GA GitHub 08/2024 – 05/2025 Vice President

● Designed, soldered a custom PCB flight computer with ESP32, BMP280, GT-U7, MPU 2080, SD card, and a Camera with SPI, I2C, and UART communication protocol in Fusion360

● Wrote C++ software for main flight program and event detection, along with library files for BMP280 pressure sensor, EBYTE LoRa radio and GT-U7 GPS

● Developed a real-time guidance system combining custom particle filter state estimation with sensor fusion for localization with Model Predictive Control/Reinforcement Learning algorithms to optimize dual-thrust rocket trajectories in C++ and Python

● Led development of GNC, earning 1st place in undergraduate software research competition among 50 participants. PROJECTS

EMG-Driven 9-DoF Prosthetic & Exoskeleton Foundation Model Python, PyTorch, SconeGym GitHub

● Processed, cleaned, normalized, and formatted a heterogeneous biomedical meta-dataset merging 9 distinct biomechanical sources, scaling to 35.3M tokens (13 EMG channels, 9 DoFs) for HPC training

● Developed an EMG-to-impedance Imitation Learning and SAC RL pipeline to train a Foundation Model to control a 9-DoF robotic prosthetic, predicting next-state kinematics and impedance parameters for torque through MoE heads

● Finetuned the model through SAC on SconeGym for custom isometric and single/dual transfemoral leg environments, integrating extensive parameter randomization to guarantee robust sim-to-real transfer Vision Language Action Model Python, PyTorch, AWS GitHub

● Fine-tuned an instance segmentation Mask-RCNN model on RoboSeg data for robot manipulator recognition achieving 96% eval score

● Created a GPT architecture from scratch in PyTorch for multi-task proprio manipulator, control, trained and validated on 700GB in AWS

Contact this candidate