About Me

An AI Engineer specializing in Speech and Natural Language Processing (NLP) with a strong background in building and deploying advanced speech recognition and natural language solutions. I am passionate about leading deep learning initiatives, transforming cutting-edge research into scalable, production-grade AI systems. I emphasize innovation, user-centric development, and operational efficiency to deliver impactful real-world applications.

🎓 Education

Master’s Degree in Computer Science (2022 – 2024) Hanoi University of Science and Technology | Good Degree, TOEIC 785

Engineer’s Degree in Software Engineering (2016 – 2021) Hanoi University of Science and Technology | Valedictorian, CPA 3.75

💼 Work Experience

Veriserve (subsidiary of SCSK Corporation)

Senior LLM Engineer

Participating in key customer projects, leveraging large language models (LLMs) to build real-world products. With designing solutions, fine-tuning models, optimizing performance, and ensuring that the products meet business requirements while remaining scalable and practical.

CMC Corporation

Head of Speech Research

Led a 5-member Speech Research Team, covering domains such as Speech-To-Text, Speaker Diarization, Speaker Verification, Voice Cloning, Voice Conversion and Audio LLM.

Vbee JSC

AI Engineer

Led the TTS (Text-to-Speech) core team in researching and deploying cutting-edge TTS models on cloud infrastructures (AWS, GCP). Developed TTS core modules, represented Vbee at major conferences and events, and enhanced the company’s reputation in the AI Voice industry.

🚀 Key Projects

AI Voice Recording System

• Headed the development of an AI Voice System for insurance partners, utilizing advanced speech recognition technologies to deliver real-time, accurate speech-to-text conversion.
• Designed and implemented speaker diarization to effectively differentiate and identify multiple speakers, improving the clarity and accuracy of transcriptions.
• Developed and integrated intent recognition models to analyze spoken content, ensuring accurate identification of user intentions.
• Optimized the system to ensure advisory content strictly adheres to prescribed scripts, improving compliance with industry standards and enhancing customer service efficiency.
• Leveraged ONNX for converting speech models and implemented OpenVINO optimizations to accelerate AI inference on Intel hardware, significantly boosting 2x model performance.

AI Call Center Caching System

• Developed for Vbee AI Call Center (AICC – vbee.ai/aicall), an AI-powered platform designed to automate customer service interactions.
• Designed and implemented a caching system leveraging the Viterbi algorithm to identify and reuse the best-synthesized audio, significantly reducing infra costs.
• Led research, implementation, and deployment on AWS, ensuring high scalability, availability, and efficient resource utilization. Achieved a stable system operation, resulting in an 85% reduction in infra costs.
• Optimized real-time model performance using asynchronous programming techniques and parallel execution to handle multiple concurrent requests efficiently.

Text To Speech Core System

• Designed and built the TTS Core Module for Vbee (vbee.vn), specializing in high-quality voice solutions.
• Led the TTS Core team in researching, developing, and deploying state-of-the-art TTS models to AWS cloud infrastructure.
• Developed a hate speech detection model integrated into a TTS system to filter harmful content before conversion to speech.
• Enabled Vbee to strengthen its position as one of the top AI voice service providers in Vietnam through this core system.

🏆 Honors & Awards