Embodied AI Engineer Humanoid Robotics Candidate

**********.*****@*****.*** 076******* Github

Embodied AI Engineer Intern – VNTT

Research and develop Vision Language Action (VLA) and Whole-Body Control (WBC) model for humanoid robots (Groot-N1, SONIC, TWIST2) combining perception, language and behavior. Designed and deployed a Digital Twin warehouse system integrating simulated environments, sensor data, and real-world inputs to enable monitoring, analysis, and intelligent decision-making. Simulated AMRs within the Digital Twin environment, processed multi-sensor data streams, and integrated ROS2-based communication for real-time control and state monitoring. Computer Vision Engineer Intern – VNTT

Developed an identity recognition system for building camera applications, including face detection, face recognition, and anti-spoofing mechanisms to enhance security and prevent fraudulent access. Accurate 6D object pose estimation for complex pick-and-place and assembly operations. Developed a voice interaction system for AMR robots, integrating speech recognition (ASR) and text- to-speech (TTS).

Ho Chi Minh City University of Technology and Engineering (HCMUTE) Bachelor of Mechatronics Engineering

GPA: 3.05/4.0

English: TOEIC 680

Chinese: HSK 3

Programming Languages: Python, C/C++, MATLAB,

Simulation & Robotics Platforms: Isaac (Sim/Lab), Omniverse, Mujoco, Gazebo Frameworks & Libraries: PyTorch, TensorFlow, Transformers, Lightning, TensorRT Vision & Embodied AI: Computer Vision (CV), Vision Language Models (VLMs), Vision Language Action (VLA), World Foundation Models (WFMs), LLM, TTS Robotics & Middleware: ROS2, Sensor Integration (IMU, LiDAR, Camera), Actuators Tools & Deployment: Git, Linux, Docker, FastAPI

Soft Skills: Technical reading skills, Research, Problem solving, Teamwork AI Engineer with more than 1.5 years of experience in end-to-end development of AI and Embodied systems, including model design, training, and real-world deployment. With a background in Mechatronics and Robotics, I bring expertise in Computer Vision, Embodied Intelligence, and Physics AI to build scalable and robust automation solutions. Sep 2021 – Nov 2025

SKILLS

SUMMARY

EDUCATION

WORK EXPERIENCE

NGUYEN MINH KHOA

AI ENGINEER

Mar 2026 – May 2026

Dec 2025 – Feb 2026

PROJECTS

2026 - Face Identification & eKYC System

Role: AI / Computer Vision Engineer

Tech: PyTorch, YOLO, RetinaFace, InsightFace, FAISS, FastAPI, OCR, WebSocket, Python Highlights:

Developed an end-to-end face identification and eKYC system integrating face detection, recognition, anti-spoofing, and OCR pipelines.

Built real-time verification and liveness detection APIs using FastAPI and WebSocket for continuous camera-based authentication.

2026 - Humanoid Robot Control System (RL & Whole-Body Control) Role: Project owner - AI / Robotics Engineer

Tech: PyTorch, Reinforcement Learning, ROS2, Isaac (Sim/Lab), OmniGraph, Python Highlights:

Developed a humanoid control framework integrating RL policies and Whole-Body Control (WBC) for coordinated locomotion and manipulation.

Built a simulation-to-deployment pipeline using Isaac Sim and ROS2, enabling real-time sensor integration (IMU, joint states) and control execution, with integrated AI models (e.g., computer vision and Vision-Language-Action) for perception and decision-making. 2026 - Digital Twin Warehouse Simulation System

Role: AI / Robotics Simulation Engineer

Tech: Isaac (Sim/Lab), Omniverse, ROS2, OmniGraph, OpenUSD, NVIDIA cuOpt, Cosmos, Python Highlights:

Built a Digital Twin pipeline integrating ROS2 communication, sensor streams (camera, IMU, LiDAR), robot navigation, and AI-based perception workflows. Integrated Cosmos Reason for warehouse scene understanding, analyzing multi-camera feeds and robot state data to generate automated status reports and anomaly alerts. 2025 - Research and Fabrication of a Delta Robot for Food Packaging Applications Role: Team Lead – AI & Computer Vision / Control Systems Tech: PyTorch, YOLO, ByteTrack, SAM2, Python, Mitsubishi Q-series PLC, TCP/IP, MATLAB Highlights:

Built a vision pipeline (YOLO + ByteTrack) for real-time detection & tracking (>30 FPS, >99% accuracy), streaming object coordinates for robot control. Developed an automated labeling system using SAM2 with >95% accuracy, enabling scalable dataset generation for training.

Contact this candidate