About Me

Long Hoa Chung at NSW Parliament House, Science Week 2023

At the NSW Parliament House during Science Week 2023, showcasing my image of human liver tissue captured by mass spectrometry—analysed from over 20,000 data points.

My journey began in biomedical research, advancing lipidomics, proteomics, and mass-spectrometry imaging. I've since expanded into AI and data analytics across finance, retail, and healthcare. Each project reflects my passion for blending science, technology, and creativity to deliver data-driven insights and solve real-world challenges.

Portfolio

I bring a unique blend of expertise spanning biomedical research and modern data analytics. From pioneering DESI-MS imaging (laboratory + instrument-acquired big data) to building AI-driven models for financial, retail, and healthcare datasets, my work demonstrates how rigorous science and data innovation can deliver impact across industries.

Project Experience

AI & Analytics (2024–2025): Optimised AI models for ethical outputs; delivered EDA, dashboards, and predictive models across finance and retail.
Healthcare & Biomedical Research (2019–2024): Built data pipelines; advanced lipidomics, proteomics, and DESI-MS imaging.
Leadership & Operations: Directed large-scale research operations and cross-functional projects.

Falcon 9 Launch SpaceX – Data Science Capstone Project

End-to-end data science project analysing SpaceX Falcon 9 launches, combining API data, exploratory data analysis, and machine learning to predict landing success.

Solution

Delivered an end-to-end pipeline integrating API ingestion, feature engineering, and machine-learning models to predict first-stage landing success.

Weather Forecast – Global Cities

Interactive global weather dashboard using live APIs to compare multi-day forecasts across major cities with robust data parsing and visualisation.

Solution

Built an API-driven system aggregating real-time weather data with user-requested city expansion via an in-app contact form.

Loan Default Prediction (Top 92%)

Developed machine-learning models to predict loan defaults using financial data, ranking in the top 92% globally in the IBM Data Science challenge.

Solution

Applied feature engineering and classification models to identify at-risk borrowers, stress-testing performance across ordinary and extreme cases.

Customer Churn Prediction

Predicted customer churn for subscription-based services using logistic regression, random forest, and gradient boosting models.

Solution

Compared models using cross-validation and AUC metrics to select robust, business-ready predictors.

IBM Data Analyst Capstone – Stack Overflow Survey

Analysed Stack Overflow developer survey data and built IBM Cognos dashboards revealing technology trends and workforce insights.

Solution

Delivered interactive dashboards highlighting emerging skills, hiring demand, and strategic workforce implications.

Skills

Data Analytics & Statistics: Turning complex datasets into actionable insights through exploratory data analysis (EDA), statistical reasoning, trend detection, and business-driven storytelling. Experienced in hypothesis testing, feature interpretation, and translating analytical outcomes into clear, decision-support narratives for stakeholders.

Programming & Tools: Proficient across Python, R, and SQL for data wrangling, analysis, and model evaluation. Experienced in building interactive dashboards and reporting layers using IBM Cognos, Looker, and Power BI, enabling transparent data exploration, KPI monitoring, and insight communication across technical and non-technical audiences.

Elastic Stack (Elasticsearch & Kibana): Architected and built Kibana dashboards for real-time monitoring, KPI tracking, and operational reporting, turning raw log and event data into clear, actionable visualisations. Configured Kibana alerting to proactively surface anomalies and threshold breaches, and designed Elasticsearch index mappings and field types to ensure accurate search, efficient querying, and reliable data structure.

Illustration of a Kibana operations dashboard with a bar chart panel, a region map built from geo_point index mappings, and a latency chart with an alert threshold breach — Kibana dashboards, maps, and alerting — architected on Elasticsearch index mappings

Machine Learning & AI: Designing and evaluating predictive and classification models with a strong emphasis on reliability, interpretability, and real-world impact. Skilled in model validation, performance metrics, bias awareness, and aligning ML outputs with operational and business constraints.

AI Training, Prompt Engineering & Human Feedback Systems: Experienced in AI training and evaluation pipelines using rubric-based frameworks, prompt engineering, and human-in-the-loop feedback. Contributed to SFT and RLHF workflows through structured annotation, response evaluation, and failure-mode analysis to improve model quality and alignment.

Soft Skills & Leadership: Strong stakeholder engagement, liaising confidently with executives, regulators, clients, and research collaborators, and simplifying technical findings for non-technical audiences. Experienced in leading cross-functional teams and multi-year projects end-to-end, from planning and resourcing through to compliance reporting and risk management, with a track record of mentoring junior staff and directing operations under pressure.

Experience

A journey from scientific discovery to data innovation — bridging research, analytics, and AI leadership.

Research Scientist

PhD scientist with expertise in lipidomics, proteomics, and DESI imaging, publishing novel findings and earning multiple research awards.

Data Scientist

Transitioned into data science and AI, delivering predictive analytics pipelines, business dashboards, and ranking Top-92 in a global Coursera challenge.

AI & Leadership

Current work spans AI model evaluation (Outlier AI), financial/operational analytics, and leadership roles integrating data strategy with ethical AI.

Publications

My research integrates big data in genomics, proteomics, and lipidomics with advanced computational and instrumental approaches. This combination has revealed new disease mechanisms in metabolic disorders, cancer, and neurodegeneration — supporting better diagnostics and healthier ageing. Key contributions include AI-driven proteomics pipelines and real-time lipidomics imaging.

Here is a selection of highlighted papers from my research in lipidomics, proteomics, and imaging mass spectrometry. Please find the full list of publications on my profile.

View my ORCID profile

ReTimeML: a retention time predictor that supports the LC–MS/MS analysis of sphingolipids

Scientific Report, 2024

I collaborated with other researchers to employ build a model that can predict the behaviours of Lipids inside a mass-spec instrument, improving accuracy and detection. The tool helps other researchers quickly reference and identify appropriate methods for their studies.

Read Paper

Deep serum lipidomics identifies evaluative and predictive biomarkers for individualized glycemic responses following low-energy diet-induced weight loss: a PREVention of diabetes through lifestyle Intervention and population studies in Europe and around the World (PREVIEW) substudy

Nature Metabolism, 2022

I contributed to a global clinical study by providing analytical guidance and lipidomics methods, connecting omics data with human samples to reveal how lipids influence the dynamics of Type 2 diabetes, its prevention, and management.

Read Paper

Ablation of sphingosine kinase 2 suppresses fatty liver-associated hepatocellular carcinoma via downregulation of ceramide transfer protein

Journal of Oncogenesis, 2022

I applied advanced data analytics to uncover how changes in liver fat metabolism contribute to non-alcoholic liver cancer and developed a novel imaging approach using digital analysis rather than the traditional microscopy. This method provided clearer visual and quantitative insights directly on real huamn tissues to support early diagnosis and treatment planning.

Read Paper