Computer Science Backend Developer

PHAM NHU THINH

Backend Developer

Dob: **/**/****

Gender: Male

Phone: 088*-***-***

Email: *************@*****.***

Address: ***/**/*** ****** *** Cu, Ninh Kieu District, Can Tho City, Vietnam PROFILE SUMMARY

A Computer Science student with a foundation knowledge in backend development, with a strong focus at building optimized, scalable solutions. Highly values self learning and collaborative teamwork. EDUCATION:

CAN THO UNIVERSITY Can Tho

Major: Computer Science 2021 – Present (Final semester) GPA: 3.2/4

SKILLS

Backend: Node.js, Express.js, JWT, Microservices, Socket.IO. Frontend: Typescript, SCSS, Javascript, Bootstrap, HTML/CSS. Databases: MySQL, Redis, MongoDB.

Deployment/Cloud Services: Docker, Git, iNET Cloud Services. PROJECTS

E-COMMERCE WEB APPLICATION Mar 2024 - Present

A microservice-based platform that enables sellers to list products, buyers to search and place orders, and administrators to manage the system with clear, role-based permissions. GitHub: https://github.com/CodeEzforme/product-management-NLCS Website: https://www.chaluatungloan.food

Responsibilities: (Personal project)

● Utilized Node.js with Pug for server-side rendering, significantly improving page load performance and SEO.

● Implemented real-time 1-to-1 and group chat, as well as order status notifications, using Socket.IO.

● Integrated a VQA-based product advisory chatbot, leveraging Vision Transformer (ViT) for image analysis and PhoBERT for Vietnamese language understanding, enabling customers to ask product-related questions and receive automated responses.

● Deployed to iNET Cloud with cluster autoscaling, health checks, custom domain configuration, and automated SSL provisioning via Let’s Encrypt.

● Leveraged Redis for session storage and caching of product listing pages to reduce database load.

● Containerized services using Docker for scalable and efficient deployment.

● Established an automated CI/CD pipeline triggered by Git pushes (build Docker images deploy to iNET Cloud), with support for rapid rollbacks.

AI-POWERED PRODUCT ADVISORY CHATBOT Jan 2025 - Present An AI microservice integrated into the e-commerce platform that analyzes product images and Vietnamese text queries to provide instant, context-aware advice—accelerating customer support, easing the burden on human agents, and reducing maintenance costs.

GitHub: https://github.com/CodeEzforme/product-chatbox-LV Website: https://www.chaluatungloan.food

Responsibilities: (Personal project)

● Collected product images by crawling URLs with Python scripts and supplemented the dataset with real-world samples.

● Preprocessed data using pandas for CSV parsing, OpenCV & Pillow for image resizing/normalization, Albumentations for augmentation, and PhoBERT for question tokenization.

● Trained a PyTorch VQA model by combining ViT image embeddings and PhoBERT text embeddings in an MLP architecture, optimized with Adam and early stopping.

● Evaluated model performance with cross-entropy loss and accuracy on validation/test sets, visualizing results using Matplotlib.

● Developed a FastAPI microservice to accept multipart form data (image + question), perform inference, and return JSON responses.

● Containerized the chatbot service with Docker to ensure consistent deployment environments and enable easy rollbacks.

● Leveraged Redis to cache inference results and session data, reducing redundant model calls and improving throughput.

Contact this candidate