I am an AI Engineer with 5 years of experience in Computer Vision and 2 years in NLP and Large Language Models, holding a Master’s degree in Artificial Intelligence. I specialize in solving business problems across multiple AI domains, including vision, language, and time-series, and deploying solutions to production environments. I work effectively both independently and as part of a team, with a strong passion for continuous learning.

• Developed novel Generative AI pipeline combining LISA (Vision-Language Model) for automated object segmentation and Stable Diffusion fine-tuned with DreamBooth for synthetic dataset generation, achieving 50% improvement in data quality for wildlife detection models.
• Reduced ecological dataset collection costs by leveraging generative data augmentation to overcome camera-trap variability in conservation applications.
• Developed and optimized YOLO models for wildlife detection, enhancing dataset diversity and supporting conservation-focused AI applications in Australia.

• Developed a multi-agent AI tutoring system that guides users through solving LeetCode/HackerRank problems step-by-step, providing idea generation, code review, and optimisation feedback rather than direct answers.
• Engineered a multi-modal workflow that supports text and screenshot inputs, built a crawler to ingest problems into MongoDB, and developed a recommendation pipeline that selects problems based on user preferences such as topic and difficulty.
• Delivered a streamlined Streamlit interface and implemented user-level analytics to track progress, visualise solved problems, and personalise each user’s learning experience.

• Delivered on-premises chatbots for customers across multiple domains, enabling reliable and scalable enterprise solutions with domain-specific knowledge integration. Tech stack: Vector databases (MongoDB, ElasticSearch), LLMs (Llama-2-13B), vLLM (serving), LangChain/LangGraph (workflow orchestration), Retrieval-Augmented Generation, Prompt Engineering
• Implemented ML models supporting Intelligent Transport System features including traffic violation detection and electronic toll collection systems deployed widely across Vietnam's highway network. Tech stack: YOLO, SAM, OpenCV, TensorRT, ONNX.
• Deployed a Social Listening system that crawls social network data and applies LLMs to deliver insights on keyword trends, brand reputation, and customer campaign reactions. Delivered an interactive dashboard for real-time analysis and monitoring.

• Led team of 4 engineers to develop Deep Learning solution for automated defect detection in mobile display manufacturing, replacing manual inspection process.
• Deployed Transfer Learning-based UNet and YOLO models for marked region detection/segmentation on edge devices, achieving 95% detection accuracy.
• Reduced manufacturing costs by 80% through minimizing false positive rates and eliminating manual quality control labor.
• Optimized models for real-time edge deployment using TensorRT and embedded systems integration.

• Integrated SAM2 with Ultralytics for automatic frame annotation, training YOLO11 for player and ball detection, and applied ByteTrack for multi-object tracking.
• Developed additional OCR models for recognition of player jersey numbers, enhancing automated game analytics.
• Developed post-analysis features including automatic player-focused camera generation, player heatmaps, and performance statistics to support data-driven insights for coaches and analysts. Tech Stack: YOLO, ByteTrack, SAM2, Ultralytics

• Trained a YOLO model on custom product datasets for accurate product detection.
• Developed real-time monitoring features: alerts for mis-placed products, low-stock/out-of-stock warnings.
• Optimized models using TensorRT 8-bit quantization for edge devices and deployed them with Triton Inference Server for fast, scalable inference. Tech Stack: YOLO, FAISS, TensorRT, Triton Inference Server

• Built comprehensive traffic monitoring system with vehicle detection, tracking, and attribute extraction (class, color, direction, license plate).
• Implemented DeepSORT for multi-object tracking and OCR for license plate recognition with 88% accuracy.
• Developed search functionality enabling queries by vehicle attributes and automated video summarization reducing review time by 75%. Tech Stack: YOLO, DeepSORT, OCR

• 📜 DeepLearning.AI Certificate: Agentic AI, AI Agents in LangGraph.
• 📜 Microsoft Azure AI Fundamentals (AI-900)
• 📜 Microsoft Azure AI Engineer Associate (AI-102)

• 🔍 Advanced NLP Projects: Applying large language models to solve real-world problems and creating AI agents
• 📊 Dashboard Development: Creating interactive visualizations with Power BI and Python
• 🤖 AI Model Deployment: Learning MLOps and model deployment best practices
• 📚 Continuous Learning: Staying updated with latest AI/ML research and techniques

I'm always interested in collaborating on data science and AI projects, especially those with social impact. Feel free to reach out!

⭐ Fun Fact: I speak English, Vietnamese, and Korean, and I love exploring how AI can bridge language barriers and create more inclusive technology solutions!