le thang

I’m a C++ Software Engineer with over 5 years of experience in AI & Embedded Systems development, optimizing Computer Vision algorithms on GPU/Edge Devices. I build real-time, high-performance video analytics that turn client requirements into reliable products.

At NAIZ CO., INC, I developed and optimized a high-performance C++ backend for CCTV, handling 1,000+ cameras with real-time monitoring and AI-powered event detection. I researched and built Multi-Camera Multiple People Tracking using Re-Identification, vector clustering, and database ranking—improving search speed 3x faster and achieving 92% accuracy Rank1 and 97% Rank10.

Previously, at ASILLA VIETNAM, I developed backend systems for AI Video Analysis and reduced latency by 60% using GStream and DeepStream pipelines across hundreds of cameras. I refactored modern C++ pipelines to avoid copying and locking, improving performance by >30%, and optimized models for GPUs and AI accelerators like NVIDIA Jetson, Qualcomm, and Hailo.

At BKAV CORPORATION, I helped engineer “AI View” for IP and AI cameras and led an AI backend team of five, owning system architecture, API development, and component management. I enjoy translating requirements into scalable designs—then profiling, tuning, and optimizing until the system runs fast, stable, and edge-ready.