Optimizing Driver Matching to Under 5 Seconds
Exploring the integration of KDTree spatial indexing and ThreadPoolExecutor to handle massive concurrency in ride-hailing systems.
I build high-performance systems and integrate AI to solve real-world problems. 236+ GitHub contributions in the last year.
I am a passionate Fullstack Developer and AI Integrator who graduated from the University of Information Technology (UIT - VNUHCM) with an Excellent GPA of 9.1/10.0.
With over a year of professional experience and two years of deep project development, I specialize in building secure, scalable architectures and bridging the gap between cutting-edge Machine Learning models and robust enterprise applications.
const developer = {
name: 'Phan Nguyen Tra Giang',
role: 'Fullstack AI Engineer',
education: 'UIT - VNUHCM',
gpa: 9.1,
skills: ['React', 'FastAPI', 'Llama-3', ...],
passion: () => {
return 'Building scalable AI systems';
}
};
Backend Developer
Matching drivers in < 5 seconds using KDTree and ThreadPoolExecutor.
Backend Developer
Dynamic pricing for 10,000 trips in < 3s based on supply/demand.
Machine Learning Engineer
Fine-tuned RoBERTa achieving 98.97% ROC-AUC with mixed precision.
Fullstack AI Developer
Multi-Agent Chatbot, AI Pronunciation Scoring, Core Dashboard.
Exploring the integration of KDTree spatial indexing and ThreadPoolExecutor to handle massive concurrency in ride-hailing systems.
A deep dive into building scalable Retrieval-Augmented Generation systems using MongoDB Vector Search and open-source LLMs.
How implementing FP16 mixed precision can drastically reduce training time and memory footprint without sacrificing accuracy.
I bring a mix of engineering depth, product ownership, and AI integration experience that helps teams move from prototype to dependable delivery.
Product + Engineering
My edge is not just writing code fast. It is understanding the tradeoff between architecture, delivery speed, maintainability, and user value. That is what helps me move comfortably between backend systems, AI features, and full product implementation.
Strength
Backend Architecture
Focus
AI Product Integration
Style
Ownership Mindset
I design APIs and system flows with performance, maintainability, and real production constraints in mind.
I do not just connect models. I shape AI features so they are usable, measurable, and valuable inside real products.
I ramp up quickly, take ownership seriously, and keep projects moving without needing constant hand-holding.
I communicate clearly with teammates, translate technical tradeoffs well, and enjoy building alongside others.
I care about finishing the right thing well, not just pushing code. That means cleaner execution, stronger follow-through, and fewer loose ends.
I can move between backend, AI workflows, and product-facing implementation, which makes me useful in fast-moving teams.
A few words that reflect how I show up as a teammate, collaborator, and builder.
University Teammate
Giang is the kind of teammate who quietly solves the hard parts. She thinks deeply about architecture, moves fast without becoming careless, and always keeps the long-term maintainability of the system in view. Working with her feels reliable because she brings both speed and structure.
I'm always interested in hearing about new opportunities and interesting projects.
Let's Connect
I'm open to backend, full-stack, and AI integration opportunities. If you have a role, freelance project, or collaboration idea in mind, feel free to reach out.
pntg2903@gmail.com
phan-nguyen-tra-giang
Phone
0941 410 532
Location
Ho Chi Minh City, Vietnam