Kietnehi.
Loading experience…
👋 Welcome to my portfolio

Hi, I'm
Phu Kiet

I am an IT student at Saigon University (SGU). I have a deep passion for Machine Learning, specifically Deep Learning for Computer Vision. I enjoy exploring how machines understand and process visual data just like humans do.

Follow me:
Portfolio Preview
Kietnehi Avatar

Focus

AI / ML

Student at

SGU

Open Source Activity

GitHub Statistics

Overview of GitHub activity, contribution streaks, and most used languages.

General Stats

Stars, commits, PRs, issues, and more

Live
Kietnehi's GitHub Stats

Quick Stats

Key metrics & achievements

Active

15+

Public Repos

AI/ML

Primary Focus

2025

Active Year

5+

Languages

Most Used Languages

Top languages based on repository statistics

Live
Top Languages

🏆 GitHub Achievements

Trophies earned from GitHub activities

Earned
GitHub Trophies

📊 Contribution Activity Graph

1,025+ contributions in the last year

Live
GitHub Activity Graph
Tech Stack

Skills & Technologies

Lang & Front

Python Python
Java Java
C++ C++
C# C#
JavaScript JS
HTML5 HTML5
CSS3 CSS
React React

AI & Data

PyTorch PyTorch
TensorFlow TensorFlow
NumPy NumPy
Spark Spark
CUDA
CUDA
Hadoop Hadoop

Back & DB

Node.js Node.js
Flask Flask
MongoDB MongoDB
MySQL MySQL
SQL Server SQL Server
PostgreSQL PostgreSQL

DevOps

AWS AWS
Docker Docker
Kubernetes K8s
Git Git
Grafana Grafana
Prometheus Prometheus
Portfolio

All Projects

A collection of my work in AI/ML, Web Development, and University Projects.

View GitHub Profile
Object Detection
Python Python OpenCV OpenCV YOLO YOLO

Object Detection Using YOLO

Object detection project using YOLO (You Only Look Once) algorithm, applied in Computer Vision tasks with OpenCV.

Testing Documentation & CI/CD
Testing Docker CI/CD Vercel / Railway Docs

E‑commerce Website — Testing & DevOps

Comprehensive software testing deliverables for the SGU clothing e‑commerce project: Dockerized app, CI/CD pipelines, deployments to Vercel & Railway. Includes test artifacts (DOCX, Excel test cases), and full architecture & behavior diagrams (C4, Activity, Sequence, Deployment, DFD, ERD). Test suites cover black‑box and white‑box approaches.

Machine Translation EN-FR
NLP NLP Translation MT Docker Docker Flask Flask

EN → FR Machine Translation

A simple web demo for English→French translation using an open-source model. Containerized with Docker for easy deployment.

Image Classification
TensorFlow Deep Learning CNN CNN Python Python

Image Classification

Building Convolutional Neural Networks (CNNs) using TensorFlow and Keras to classify images into different categories with high accuracy.

Fake News
NLP NLP Sklearn Sklearn Python Python

Fake News Detection

System using Natural Language Processing (NLP) and machine learning with Scikit-learn to detect and classify fake news articles.

Housing Price
Regression Regression Pandas Analysis Python Python

Housing Price Prediction

Housing price prediction model based on dataset attributes, utilizing regression algorithms for accurate estimations.

Translation Generation
Python Python NLP NLP PyTorch PyTorch Hugging Face Hugging Face

Translation & Text Generation

Advanced NLP project for multilingual translation and text generation using Transformer models. Supports multiple languages and includes both training and inference pipelines.

Transformer Multilingual Deep Learning
Web Scraping
Selenium Selenium Data Mining Data Mining Python Python

Web Scraping & Prediction

Automated tools to scrape data from websites using Selenium and build predictive models based on the collected data.

Fabric Agency
Web Dev Web Dev Project Project PHP PHP JS JavaScript

Fabric Agency Website

A website project designed for a fabric agency, showcasing products and agency information.

Copilot Visual Online
Gemini Gemini AI AI Assistant Vite Vite Monaco Web App

Copilot Visual Online

Visual Code Online is a browser-based code editor powered by Monaco and Vite, offering live preview, multi-file project management and auto-save. It includes an AI Code Copilot that indexes the repository with Jina embeddings, performs semantic retrieval with Faiss, and uses Google Gemini to generate context-aware answers about the code.

Face Recognition Web
OpenCV Computer Vision Face Recognition PyTorch PyTorch Web Web App

Face Recognition Web

A web-based face recognition system using deep learning models to identify and verify faces in real-time. Developed with PyTorch for model inference and OpenCV for image processing.

[Image of face recognition technology workflow using PyTorch]
FOODFAST
GitHub GitHub AWS AWS K8S K8S Vercel Vercel Railway Railway Prometheus Prometheus Grafana Grafana CI/CD CI/CD

FOODFAST Application

A food ordering application project. Developed in collaboration with team members.

RAG Multimodal LLM
RAG Ollama LLM HuggingFace HuggingFace Docker Docker

RAG & Multimodal LLM System

Built a Retrieval-Augmented Generation (RAG) system using open-source Large Language Models from Hugging Face such as Ollama, Qwen, Mistral, and Tiny LLM. Integrated BLIP for Visual Question Answering (VQA) and image captioning tasks, with the entire system containerized using Docker for scalable deployment.

AI Agent
Gemini Gemini Meta Meta Microsoft Qwen Qwen Python Python

AI Agent System

An intelligent AI Agent capable of performing calculations, plotting graphs, analyzing data, and searching for the latest information. The agent leverages models from Google, Meta, Microsoft, and Alibaba.

Competitions

Competitions & Hackathons

Participated in various AI/ML competitions and hackathons, applying cutting-edge technologies to solve real-world problems.

VNPT Hackathon - Age of Unicorn
Computer Vision OCR LLM AI Agent Audio Recording

AI-Powered Meeting Assistant

Developed an intelligent meeting assistant platform leveraging Computer Vision for participant recognition, OCR for document processing, Large Language Models for content analysis, and AI Agents for automated task management. The system includes audio recording capabilities to capture and transcribe meeting discussions in real-time.

Status: Submitted proposal - Did not advance to final round

Cursor Hackathon - UAV Danger Detection
Computer Vision Object Detection UAV Emergency Response

UAV-Based Flood Rescue Detection

Developed an autonomous UAV (Unmanned Aerial Vehicle) system for detecting people in danger during flood disasters. Using advanced computer vision and object detection algorithms, the system can identify individuals trapped in flood zones from aerial imagery, enabling faster emergency response and rescue operations. This project combines drone technology with AI to save lives in critical situations.

Impact: Real-time detection system for emergency disaster response

Available for collaboration

Let’s build something awesome

Bạn có cùng đam mê AI/ML hoặc muốn trao đổi về công nghệ? Nhắn mình một câu là mình phản hồi liền.