Hi, I am Soon Hao

I created this website to showcase my projects and experience, also mostly just because I was bored and wanted a fun build. I'm always looking for ways to upskill myself, especially when it comes to tech, data, and AI.

I am a massive sports fan, particularly with Manchester United and LeBron James. If you ever want to debate match results, stats, or basketball takes, I'm always down to chat.

I also love exploring random cities when I travel, so I'm planning to use this site as a little corner of the internet to post and share more of my travel photos, too.

Work & Internship Experience

Digital Solutions Analyst

ST Engineering

Current

Business Architecture Analyst

Accenture

Oct 2024 – Sep 2025

Robotic Process Automation Intern

CPF Board

Jan 2024 – Sep 2024

Business Analyst Intern

Shopee

Aug 2023 – Dec 2023

IT Business Analyst Intern

Johnson & Johnson

Jan 2023 – Jun 2023

Data Analyst Intern

National Community Leadership Institute (NACLI)

Jun 2022 – Aug 2022

Testing & Data Analyst Intern

Tracesafe

Jun 2021 – Aug 2021

Projects Overview

In my free time, I enjoy diving into data-driven projects that blend analytics, prediction, and storytelling. Whether it is analyzing MVP trends, debating the GOAT in basketball, or putting together an analytical dashboard to track lottery numbers I love using data to explore ideas and uncover insights.

I Built an AI Expense Tracker That Lives in Telegram

I built a Telegram bot that lets me log expenses as plain-text messages, like "12.50 kopi at foodcourt." An LLM extracts the amount, category, and description, and writes it straight into a Google Sheet, no app, no forms, just a text message.

At the end of each month, I ask the bot for a review, and it compares spending across categories and tells me, in plain language, where I overspent and what's worth cutting back on. The whole thing runs on free-tier infrastructure (Google Apps Script, Gemini API), with no backend or hosting cost.

Read the full article on Medium

From Stats to Superstars: Using Data Science to predict the 2024–2025 NBA MVP

As a basketball fan and data enthusiast, I was curious whether we could predict MVP shares using players and teams season stats. This curiosity led me to build a machine learning model that forecasts MVP voting outcomes based on historical data.

I experimented with an ensemble model (Linear Regression, Gradient Boosting, XGBoost). Throughout the project, I focused on model interpretability, cross-validation, and feature importance to understand which metrics truly influence MVP results.

Read the full article on Medium

LeBron vs. Jordan: The Definitive GOAT Debate by the Numbers

The LeBron vs Jordan debate has always been filled with strong opinions, so I wanted to cut through the noise and approach the question from a more analytical angle. Using a blend of statistics, historical context, and weighted metrics, I compared the two players across career longevity, regular season dominance, playoff success, and awards.

I built this as a structured, data-first deep dive to spotlight patterns fans often overlook, like how scoring trends changed by era. Beyond just the stats, the project pushed me to think critically about framing narratives with data and storytelling.

Read the full article on Medium

Toto Dashboard: Data-Driven Analysis

This project is an interactive analytical dashboard built to explore historical Toto results and statistics. Because each draw is completely independent of the past, this tool isn't about beating the odds or promoting gambling. It is purely a fun way to dive into data, numbers, and frequency stats.

The dashboard lets you explore past draw trends and visualize data in one place, serving as a practical, lightweight example of applying data analytics to random numbers just for the sake of curiosity and good engineering.

My Travel Dump

I thought this would be a nice way to log my travels since I have a website. (Not a blogger)

Click Here To View