Data Scientist

Designing intelligent systems that turn data into decisions

Transforming complex data into production-ready ML systems. Master of Data Science, University of Melbourne.

View My Work Get in Touch

Published Springer ERCICA ML Research Paper

Client Impact Phoenix Australia Automated analytics for 190+ courses

Industry Experience Software Engineer Tata Consultancy Services

Scroll to explore

Selected Work

Projects that showcase my expertise in data science and engineering

Qualtrics API

SQLite Cache

Gemini AI → SBERT Fallback

Dash Analytics

190+ Courses

Self-Improving

Featured Project

Phoenix Australia Survey Analysis Dashboard

AI-powered automated pipeline for analyzing training evaluation surveys using the Kirkpatrick Four-Level Model. Built for Phoenix Australia, a mental health non-profit affiliated with the University of Melbourne, supporting 190+ online courses.

Multi-tier fallback architecture: Cache → Gemini AI → SBERT → Semantic similarity

Self-improving system: stores all classifications in SQLite, retrains SBERT from accumulated data

Human-in-the-loop learning: user corrections persist and improve future classifications

Question hashing for deduplication ensures instant lookup of previously seen questions

Statistical pre/post comparison with t-tests, Cohen's d effect sizes, and confidence intervals

Professional LaTeX PDF reports with sentiment analysis and theme extraction

190+ Courses Supported

3-tier Fallback System

Python Dash/Plotly Gemini AI SBERT SQLite Docker Qualtrics API LaTeX

View on GitHub

AEMO / Weather APIs

Azure Databricks (Medallion)

XGBoost / CatBoost (FLAML)

Power BI + FastAPI

26,156 Records

Live Pipeline

Featured Project

Melbourne Electricity Market Analysis

End-to-end data platform built on 3 years of Victorian electricity market data. Raw data from three public APIs flows through a medallion pipeline on Azure Databricks, powers three production ML models, and surfaces as a four-page Power BI dashboard with a live FastAPI endpoint.

Medallion architecture (Bronze → Silver → Gold) on Azure Databricks with Delta Lake

Live pipeline: hourly AEMO price ingestion, nightly silver refresh, continuous production updates

Price prediction model (XGBoost) achieving R² 0.89 with $14.90 MAE on wholesale prices

Demand forecasting (CatBoost) with 2.5% MAPE and R² 0.98 on test data

Spike classifier with 0.99 ROC-AUC using Platt calibration on 12.4:1 imbalanced classes

Summer Paradox discovery: renewable generation falls on hottest days (29% vs 55% on cool days)

R² 0.89 Price Prediction

3 Models In Production

3 Years Market Data

Azure Databricks PySpark XGBoost CatBoost FLAML Power BI FastAPI Delta Lake

View on GitHub Live API

More Projects

CouchDB Cluster (Data)

Jupyter (Analysis)

NGINX (Frontend)

Cloud Infrastructure

Australia Social Media Analytics Platform

Scalable cloud-based analytics system investigating the relationship between social media sentiment and crime rates across Victorian LGAs. Processed 61GB Twitter dataset.

Docker CouchDB Ansible NLTK

View Details

Raw ECG Data

Data Cleaning

Feature Engineering

Classification Models

ML Research

Cardiac Arrhythmia Prediction System

Published machine learning research classifying 13 cardiac conditions from ECG data. Featured in Springer ERCICA Volume 1 with 72% accuracy.

scikit-learn PCA t-SNE Research

View Details

Data Visualizations

Interactive dashboards and data stories

View in full screen for optimal experience

Accidents Analysis Dashboard

Interactive analysis of accident data patterns, trends, and contributing factors

Tableau Analytics

Click to view dashboard screenshots

Phoenix Australia Survey Dashboard

AI-powered analytics dashboard for training evaluation surveys using Kirkpatrick Model

Dash/Plotly Python

Click to open Streamlit dashboard

Cardiac Arrhythmia Dashboard

A Streamlit dashboard for visualizing and predicting cardiac arrhythmias from ECG data.

Streamlit Python

View on GitHub

Click to view interactive Power BI dashboard

Melbourne Electricity Market Dashboard

Four-page Power BI dashboard covering price intelligence, generation mix, weather correlations, and AI price predictions for Victoria's electricity market.

Power BI Analytics

View on GitHub

About

Data Scientist with a Master's degree from University of Melbourne and 2 years of software engineering experience at Tata Consultancy Services.

I'm fascinated by the intersection of AI and real-world impact—building systems that don't just crunch numbers, but genuinely help people make better decisions. Whether it's analyzing 100,000+ social media posts to uncover sentiment patterns or automating mental health training evaluations, I thrive on turning messy, complex data into clear, actionable insights.

My approach combines rigorous engineering with curiosity-driven exploration. I believe the best solutions come from deeply understanding both the data and the humans who will use it. Currently exploring AI agents and LLM applications— the frontier where automation meets intelligence.

Recent highlights: Top 10% in a Kaggle AI-generated text detection competition (84% accuracy), and building fault-tolerant cloud systems that process massive datasets with 70% faster deployment through automation.

Download Resume