Computer Science Artificial Intelligence

NGUYEN GIA HUY

SUMMARY

Computer Science graduate with a strong interest in Artificial Intelligence, especially in audio processing and speech technologies. Experienced in fine-tuning pre-trained models such as NeMo STT and mT5 for Vietnamese language tasks. Hands-on with chatbot development, regex-based text retrieval, and Selenium automation. Solid understanding of NLP, data preprocessing, and database design using MySQL. Eager to apply AI knowledge in real-world projects and continue growing in the field of AI engineering.

WORK EXPERIENCE

EDUCATION

Fine-tuned the ars nemo stt_en_fastconformer_ctc_large speech-to-text model for Vietnamese language.

Crawled and preprocessed legal text data from vbpl.vn to develop a chatbot with regex- based search functionality.

Annotated datasets and verified labeling accuracy performed by other annotators to ensure data quality.

AI Intern – CMC Telecom Aug 2024 - Present

Personal Projects

Vietnamese QA Chatbot with XLM-Roberta

Fine-tuned the xlm-roberta-base model for Vietnamese question answering. Built a chatbot capable of understanding context and answering user questions in Vietnamese. Training Center Management System (MySQL)

Designed and implemented a relational database for managing courses, classes, students, and registrations using MySQL.

Multilingual Translation with mT5

Fine-tuned the mt5_translation.pt model for customized translation tasks involving multiple languages, focusing on improved accuracy for Vietnamese-English translation. Vietnamese Speech-to-Text with Whisper

Fine-tuned OpenAI's Whisper model for Vietnamese speech recognition using the doof- ferb/infore1_25hours dataset. Improved transcription accuracy for Vietnamese audio, showcasing expertise in speech-to-text model adaptation. Bachelor of Computer Science

Hanoi of Science and Technology University Join with Troy University Program GPA: 3.2/4.0

Sep 2021 - Aug 2025

2F, Quang Trung, Tran Hung Dao, Hoan Kiem, Ha Noi • 091******* • ***********@*****.***

Contact this candidate