Wai Phu Paing | Data Science Portfolio

Selected Work 05 Projects

Building intelligent systems end‑to‑end

Focus Areas

Document AI

Computer Vision

Machine Learning

Cloud AI

Featured Capstone

Receipt Scan AI

Built an end-to-end Document AI system that converts receipt images into structured financial data using OCR, LayoutLMv3, OpenCV, token classification, geometry-aware extraction, validation rules, and LLM prompt engineering for extraction support and auto-correction.

Phase 1

OCR + Preprocessing

Phase 2

Item Extraction

Phase 3

Metadata + Validation

Python LayoutLMv3 OCR Document AI OpenCV LLM Prompting Angular

Focus area: intelligent document processing rather than broad text-language modeling. This presents the project accurately while still showing strong AI engineering skills.

Project Links

GitHub Code Demo Video

Big Data + Cloud AI

Automated Damage Assessment System for Natural Disasters

Built a cloud-based AI system for large-scale disaster damage assessment and risk prediction using deep learning and ensemble models. Integrated multi-source data pipelines and deployed scalable workflows on Azure for real-time analysis.

Azure ResNet DenseNet EfficientNet XGBoost

View on GitHub Report View Dashboard

Computer Vision

Marine Debris Detection

Developed a YOLOv8-based object detection system for marine debris classification. Applied class-wise data augmentation and targeted oversampling to reduce class imbalance, improving detection performance with mAP@0.5 from 0.56 to 0.70 and significantly boosting minority-class recall.

YOLOv8 Albumentations OpenCV mAP F1-score

View on GitHub Report

Industrial AI Project

EDM Cutting Parameter Prediction

Built a data-driven machine learning pipeline to predict EDM cutting parameters from real production data, enabling automated decision support and improving machining efficiency in industrial environments.

Python Machine Learning Data Analysis Regression

View on GitHub

Machine Learning

Crop Yield Prediction

Developed Random Forest and Linear Regression models with preprocessing, feature engineering, and evaluation using R², MAE, MSE, and RMSE.

Random Forest Regression Pandas

View on GitHub

Technical Skills

AI & Data Science

Machine learning, data preprocessing, feature engineering, model evaluation, Document AI, OCR, information extraction, data cleaning, and validation pipelines.

Libraries

NumPy, Pandas, OpenCV, PyTorch, LayoutLMv3, YOLOv8, XGBoost, OCR tools, and LLM API integration.

Software Engineering

Python, C#, TypeScript, JavaScript, PHP, SQL, .NET Core, ASP.NET, REST APIs, multithreading.

Cloud & Tools

Microsoft Azure, CI/CD, Git, Bitbucket, Jira, Agile, Scrum, SQL Server, MySQL.

Strengths

End-to-End AI

From raw data and images to model output, validation, UI, and deployment-ready pipelines.

Engineering Foundation

Strong backend, API, database, and automation experience from production software roles.

Research Mindset

Comfortable with evaluation metrics, experimentation, model comparison, and technical writing.

Professional Experience

Jun 2022 – Jan 2024

Software Engineer II

Makino Asia Pte Ltd

Singapore

Developed Angular, .NET 6, C#, REST API, and SQL web applications.
Built multithreaded .NET services for robot and CNC/PNC machine communication.
Supported 24/7 automated manufacturing systems across multiple countries.

Jul 2021 – May 2022

System Analyst

NUS Information Technology

Singapore

Built backend services with C# .NET Core, REST APIs, and SQL Server.
Developed Angular frontend features and supported Agile sprint delivery.
Handled UAT, bug fixes, performance improvements, data migration, and batch processing.

2016 – 2021

Developer Roles

Anxing Technology · Systematic Business Solution · MIT

Developed full-stack applications, APIs, ERP systems, POS systems, reporting modules, and backend services.
Worked with PHP, Laravel, MySQL, ASP.NET, C#, Java, SQL Server, JavaScript, and Elasticsearch.

Education

Master of Science in Computer Science

Data Science

University of Regina, Regina, SK, Canada | Apr 2024

Post Graduate Diploma in System Analysis

National University of Singapore | Feb 2021

Bachelor of Engineering

Computer Engineering

University of Technology (Yatanarpon Cyber City), Myanmar | Nov 2015

Research

IEEE CISCE 2026

Research + Implementation (ESRGAN Super-Resolution)

Evaluated locally discriminative learning for GAN-based super-resolution (ESRGAN), improving image quality and reducing hallucination artifacts. Implemented and tested on UC Merced dataset.

Code Paper