Aspiring AI Engineer and Data Scientist Intern

HUYNH TIEN DUNG

Phone: 039******* Location: Ho Chi Minh, Viet Nam

Email: *****************@*****.*** Linkedin: https://www.linkedin.com/in/hduxng200205 Github: https://github.com/Hduxng

A. ABOUT ME

Third-year Computer Science student at the University of Information Technology (UIT) with a deep passion for Artificial Intelligence and Data Science. I possess hands-on experience in building end-to-end systems, including Recommendation Systems (processing ~40 million data rows) and Semantic Search Engines. With a solid foundation in Deep Learning (Transformers, CNN, GenAI) and strong proficiency in Python/PyTorch, I am seeking an internship opportunity to apply my knowledge to real-world problems and contribute tangible value to the organization.

B. PROJECTS

1. Video Retrieval System – AI Challenges Competition Simulation Domain

Computer Vision, NLP, Information Retrieval.

Responsibilities

Semantic Search Engine Design: Designed a system to retrieve specific video segments via natural language queries.

Computer Vision & Multimodal Learning: Implemented Keyframe Extraction and utilized BEiT-3 to generate multi-modal embeddings for video frames and text queries.

Vector Database Management: Constructed a scalable vector database using Milvus for high- dimensional indexing and efficient similarity search.

Data Pipeline Optimization: Optimized the processing pipeline to successfully handle and index over 100GB of video data.

2. Retail Recommendation System (Concung.com Store) Domain

Recommender Systems, Big Data Processing, Retail Analytics.

Responsibilities

System Architecture: Architected a 2-stage Recommender System processing ~40M transaction rows, utilizing Item-Item CF for retrieval and handling cold-start scenarios using Global Trending strategies.

Model Engineering: Engineered a Hybrid Ranking Model ensembling LightGBM (0.6) and DeepFM (0.4) to capture both numerical statistics and high-order feature interactions, optimizing for Precision@10.

Feature Engineering: Designed domain-specific features (e.g., Baby Age, Brand Loyalty) and distinct inference strategies for user segments, effectively personalizing recommendations based on customer lifecycle..

3. Fine-tuning Stable Diffusion for Dong Ho Folk Art Generation Domain

Generative AI, Computer Vision, Deep Learning.

Responsibilities

Data Preparation: Curated and preprocessed the Dong Ho Folk Art dataset (sourced from MMLab), optimizing training inputs via image resizing (512x512) and detailed captioning for better semantic alignment.

Generative AI Fine-tuning: Fine-tuned Stable Diffusion v1.5 using PEFT techniques (LoRA, Textual Inversion, IP-Adapter) to conduct a comparative analysis on style fidelity versus prompt adherence.

Model Evaluation: Assessed performance using CLIP Score and CSD Cosine Similarity, successfully reproducing unique artistic features like "Diep" paper texture and natural folk colors. 4. Survey and Improvement of Deep Learning Models for Tea Leaf Disease Classification Domain

Computer Vision, Deep Learning.

Responsibilities

Data Processing: Processed the teaLeafBD dataset (5,278 images) across 7 classes, implementing a custom dataset class and applying data transformations (Resize, Normalize).

Data Balancing Strategy: Addressed the class imbalance problem by implementing a WeightedRandomSampler, ensuring balanced training batches.

Data Augmentation: Applied advanced techniques including Random Resized Crop, Horizontal Flip, Rotation, Color Jitter, and Random Erasing to improve model generalization.

Model Comparison & Training: Conducted a comparative study of modern architectures

(EfficientNet, ViT, Swin Transformer), training them using PyTorch with AdamW optimizer and CrossEntropyLoss.

5. Other Projects

Predicting Bestseller Courses & Leaning Path Recommendation.

Email Phishing Detection.

Los Angles Crime Data Analysis & Warehousing.

Clustering: Gaussian Mixture Models and Expectation-Maximization method. C. SKILLS, CERTIFICATIONS & EDUCATION

1. TECHNICAL SKILLS

Languages & Core: Python (Advanced), SQL, C++, OOP, Data Structures & Algorithms.

Deep Learning: PyTorch, TensorFlow, Transformers (Swin, ViT, BEiT-3), Diffusion Models (Stable Diffusion, LoRA, IP-Adapter), CNN, LSTM.

Machine Learning: LightGBM, XGBoost, Random Forest, Scikit-learn, DeepFM, Recommender Systems, Genetic Algorithm.

Data Engineering: Polars (High-performance), Pandas, NumPy, Milvus (Vector DB), Selenium, OLAP. 2. CERTIFICATIONS

TOEIC 870 (12/2024)

APPLICATIONS OF AI FOR ANOMALY DETECTION - NVIDIA (2025)

COMPUTATIONAL THINKING - University of Michigan (2024) 3. EDUCATION

University of Information Technology (UIT) 2023 – Now GPA: 3.2 / 4.0

Major: Computer Science

Contact this candidate