An intelligent retrieval-augmented generation system that processes and queries across multiple content types including text documents, images, and videos. Features custom vector embeddings, hybrid search, and LLM-powered re-ranking for precise multimodal information retrieval.
Hi, I'm
Hieu Nguyen
AI Engineer
Building production-ready AI systems with NLP, RAG, and LLMs. Passionate about bringing AI into real-world applications.
About Me
AI Engineer with Master's degree from JAIST, specializing in production ML systems
Hello there!
I'm Hieu Nguyen, an AI Engineer from Vietnam with a passion for transforming research into production-ready systems.
My expertise lies in Natural Language Processing, RAG systems, and Large Language Models. I specialize in building scalable AI solutions that solve real-world problems.
🎓 I hold a Master's in Information Science from Japan Advanced Institute of Science and Technology (JAIST) in Japan, where I deepened my knowledge in information retrieval and machine learning.
💡 When I'm not coding, you'll find me exploring self-hosted solutions, contributing to open source, or writing about AI/ML on my blog.
Education
Master's in Information Science
Japan Advanced Institute of Science and Technology (JAIST)
Japan
Tech Stack & Skills
AI/ML
Languages
Backend/APIs
Databases
DevOps/Tools
Data & Engineering
🚀 Featured Projects
Check out some of my recent work and contributions
A comprehensive English proficiency assessment platform powered by AI. Features adaptive reading tests with LLM-generated questions, speech-to-text evaluation for speaking tests using Whisper, and detailed AI feedback on writing. Provides personalized insights on grammar, vocabulary, fluency, and coherence across all skills.
A modern open-source online judge and contest platform system for Vietnamese students. Features automated code evaluation, interactive programming practice, and 1000+ programming exercises with comprehensive solutions.
✍️ Latest Articles
Thoughts on AI, ML, and software engineering
Inverted HyDE: Solving Real-World Dense Retrieval Challenges
An innovative approach to dense retrieval that addresses practical limitations of HyDE by flipping the script - generating hypothetical queries offline instead of hypothetical documents in real-time.
Stop Using requirements.txt
The requirements.txt is a legacy dependency management tool that is no longer fit for modern Python projects. We need a better dependency management tool.
Fine-tuning Flux.1-dev LoRA on yourself
This blog serves as my personal guide to fine-tuning Flux.1-dev LoRA to generate high-quality, lifelike images of myself—all without the hassle of taking photos.
Let's Build Something Amazing Together
Looking for an AI engineer to bring your ideas to life? I'm available for consulting, collaboration, and full-time opportunities.