AI Research Engineer · NLP PhD Student · Multimodal Systems Builder
Building practical AI systems across language, vision, retrieval, and agentic workflows.
I am Vo Hoang Nhat Khang (Chris), a PhD student in Natural Language Processing at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI). I work on vision-language models, multimodal learning, uncertainty estimation, and Vietnamese NLP. My style is simple: take research ideas seriously, implement them cleanly, and turn prototypes into systems people can actually use.
| Research | Vision-Language Models · Multimodal Learning · Uncertainty Estimation · Vietnamese NLP |
| Models | LLMs · VLMs · CLIP-style models · Transformers · Retrieval-Augmented Generation |
| Engineering | Python · C++ · Java · Backend APIs · Data pipelines · Research prototyping |
| Systems | RAG applications · Agentic AI · Healthcare AI · Vietnamese language tools |
| Project | Type | Description |
|---|---|---|
| HuTieuBERT | Research | Morpheme-aware Transformer work for Vietnamese NLP, accepted to ACL 2026 Main Conference. |
| V7 | Product | Vietnamese AI input method designed for faster and more flexible typing. |
| Namizer | NLP Tool | Vietnamese tokenizer that decomposes words into five linguistic components. |
| PhoCLIP | Multimodal | CLIP-style Vietnamese vision-language model for text-image understanding. |
| TI-JEPA | Research | Text-image joint-embedding predictive architecture for multimodal fusion. |
| LumbarCLIP | Healthcare AI | Multimodal CLIP-based framework for low back pain diagnosis. |
| VitalFit Persona | Agentic AI | AI-powered healthcare application with agentic workflows. |
| HCMUT Chatbot | RAG System | Retrieval-augmented chatbot for university admission support. |
| Synthetic Generator | Data System | Synthetic dataset generation system for ML workflows. |
