Machine Learning Ai Engineer

HUYNH NGOC DUY KHUONG

AI ENGINEER

§ github.com/Ripefog ï linkedin +84-944****** # ******************@*****.*** OBJECTIVE

Driven by a strong passion for Artificial Intelligence, I have focused my studies and research on Machine Learning, Computer Vision, and related fields. With hands-on experience in developing deep learning models and proficiency in Python and PyTorch, I thrive on creating AI-powered solutions to tackle real-world challenges. My objective is to continuously refine my expertise, explore groundbreaking AI applications, and make meaningful contributions to the evolution of the field. EDUCATION

Ho Chi Minh City University of Technology September 2022 - Now Bachelor of Engineering in Computer Science Cumulative GPA: 7.1/10 Tran Hung Dao High School for The Gifted August 2019 - June 2022 English class

Key courses taken

Discrete Mathematics for Computer Science, Data Structure and Algorithms, Database Systems, Operating Systems, Computer Networks, Machine Learning, Big Data, Data Mining and Principle of Programming Language and AIO2024

EXPERIENCES

IVS JSC November 2024 - May 2025

AI/ML Intern Robotic

Developed robot control algorithms integrated with real-time Computer Vision.

Built a face detection and tracking system using TensorFlow and Mediapipe.

Integrated Automatic Speech Recognition (ASR) and Natural Language Processing

(NLP) to enable voice-based interaction with the robot.

Combined multiple modules (CV, NLP, control) into a unified robotic system using ROS

(Robot Operating System).

Technologies: ROS, Python, C++, TensorFlow, Mediapipe, OpenCV, SpeechRecognition. FPT Telecom May 2025 - Now

AI Engineer Intern

Developed AI Vision models (OCR, Object Detection) using YOLO, PaddleOCR, and CNN.

Performed data preprocessing (resize, crop) with OpenCV and trained models on the COCO dataset.

Deployed APIs with ONNX/TensorFlow Lite, optimized hyperparameters, and achieved over 80% accuracy.

Created and deployed APIs using Flask/FastAPI on servers (Render/Heroku); the API functioned correctly

Technologies: Python, PyTorch, TensorFlow, GitHub, Docker. 1

SKILLS

Programming Languages:

• Frontend: Java, JavaScript, HTML, CSS

• Backend: C/C++, Python, SQL

Machine Learning & Deep Learning:

• Proficient in CNNs, RNNs, LSTM, Transformers, GenAI, NLP

• Frameworks: PyTorch, TensorFlow, Keras, LangChain, Autogen, Hugging Face

• MLOps: CI/CD for ML, model versioning, deployment Software Development & Tools:

• Git/GitHub, Docker, Linux, FastAPI, Flask

RESEARCH EXPERIENCE

AI Lab – AI VIET NAM 2024 – Present

Research Assistant (Multimodal AI & VQA)

Conducted research on multi-modal learning and autonomous agents for Visual Question Answering (VQA) tasks.

Read and implemented key ideas from state-of-the-art models: BLIP-2, GPT-4V, MM-ReAct, MM-Vid, HuggingGPT.

Trained and fine-tuned large-scale vision-language models to generate answers and explanations on VQA-X and ViVQA datasets.

Applied explanation-aware metrics (faithfulness, plausibility) to evaluate model justifications.

Explored methods to integrate image-text reasoning, generation, and justification in agentic pipelines.

PROJECTS

Priniting Management System November 2024

Software Engineering

Designed and implemented a web page to manage a Printing system based on Javascript and GoLang

Technologies Used: Vite, React Router DOM, Docke, Makefile, Yarn, ReactJs

Programming Language: JavaScript, HTML,CSS

Bittorent Application November 2024

Computer Network

This project simulates a Peer-to-Peer (P2P) network system similar to BitTorrent, where nodes can share, search, and download files from the network.

Technologies Used: Socket, Flask

Programming Language: Python

Simple Chatbot November 2024

AIO2024

This project is a web application that mimics the functionality of ChatGPT, allowing users to interact with an AI model in a chat format. Built using Streamlit, it provides a user-friendly interface for chatting with the AI.

Technologies Used: Streamlit, FastAPI, Uvicorn, OpenAI GPT API 2

Programming Language: Python

IoT with CNN October 2024

AIO2024

This project is an intelligent IoT system that allows users to control a light bulb using only finger gestures. The system utilizes a Convolutional Neural Network (CNN) to recognize hand gestures from a camera, then sends signals to the IoT device to turn the light on/off or adjust its brightness.

Technologies Used: CNN, Pytorch, Mediapipe, CV2, etc...

Programming Language: Python

Retrieval Information from Image June 2024

HCMC Hackathon

Our hackathon project focuses on [Digital Empowerment] to enhance digital experiences. We have developed an innovative solution using an ensemble model approach. Our solution utilizes EasyOCR and image captioning models to extract insights from images. These insights are then analyzed using the large language model LLAMA3. This enables automatic detection and evaluation of various elements crucial for brand experience enhancement at HEINEKEN Vietnam, including brand logos

(Heineken, Tiger, Bia Viet, Larue, Bivina, Edelweiss, Strongbow), product identification (beer crates and bottles), customer engagement, advertising materials (posters, banners, signage), and contextual details (venues like restaurants, bars, grocery stores).

Technologies Used: EasyOCR, LLAMA3, TensorFlow/PyTorch, Ensemble Model Approac, Image Captioning Models

Programming Language: Python

CERTIFICATIONS

• Math, Programming and Data Science Foundation AI VIETNAM, 30/09/2024 Credential ID: 97827818

• Basic Deep Learning AI VIETNAM, 02/12/2024

Credential ID: 64062993

• Computer Vision and NLP AI VIETNAM, 03/06/2025

Credential ID: 40755581

• Deep Learning AI VIETNAM, 03/06/2025

Credential ID: 89848491

• GenAI and LLMs AI VIETNAM, 03/06/2025

Credential ID: 40860367

• Machine Learning AI VIETNAM, 03/06/2025

Credential ID: 15145675

3

Contact this candidate