About Me
Lim Boon Yan Edward
Aspiring Data Scientist | Analytics, Machine Learning and Business Decisioning
I am an analytics professional moving deeper into data science, with APAC experience across digital commerce, service cost forecasting, inventory optimization, and business operations. I am currently pursuing the Master of IT in Business at Singapore Management University, specializing in Data Science and Analytics, and building projects that connect business questions to data preparation, machine learning, evaluation, and decision-ready outputs.
Python/R
Machine learning, forecasting, and visual analytics
ML Pipeline
Data preparation, feature engineering, modeling, and evaluation
Forecasting
Demand, cost, and inventory analytics experience
APAC
Business analytics across regional markets
Data Science Focus
Predictive Modeling
Building forecasting, classification, clustering, and deep learning workflows that connect model outputs to business decisions.
Analytics Engineering
Preparing reliable datasets, automating recurring workflows, and structuring analysis so insights can be reused and scaled.
Business Translation
Framing open business problems into measurable analytical questions, then communicating trade-offs clearly to stakeholders.
Experience
Data Analytics Specialist, Digital and E-Commerce - APAC
HARMAN International
- Built real-time analytics dashboards for JBL eShop campaign performance and optimization insights across APAC markets.
- Designed a unified reporting template for 14+ distributors, consolidating online/offline sales, campaign, and category data into a centralized dataset.
- Automated financial reporting workflows, improving standardization and saving 8+ hours per month.
- Analyzed e-commerce and paid media data across TikTok, Shopee, Lazada, Tokopedia, Amazon, CPAS, and Meta for quarterly business reviews.
Business Data Analyst - APAC
Sony Electronics South East Asia Pacific
- Applied forecasting and risk analysis models to support budgeting decisions and repair cost provisions across APAC.
- Analyzed historical and forecasted service cost data to identify anomalies, gaps, and key cost drivers.
- Partnered with cross-functional stakeholders to define forecasting assumptions and evaluate what-if scenarios.
- Automated multi-country forecast consolidation with Excel VBA for cross-regional analysis.
Business Operations Analyst
Hewlett Packard Enterprise
- Used simulation and scenario analysis to identify inventory and asset devaluation drivers, helping mitigate losses by approximately $200K USD per quarter.
- Conducted cost-benefit analysis on defective product return strategies, reducing spare parts asset devaluation by 10%.
- Developed Power BI dashboards integrating multiple data sources for inventory monitoring and excess stock identification.
- Automated reporting workflows with Excel VBA, reducing manual effort by approximately 60%.
Business Application Support Consultant
MicroChannel Singapore Pte Ltd
- Resolved SAP Business One system and data issues across finance and sales modules.
- Gathered business requirements and translated them into system solutions and proof-of-concept implementations.
Skills
Analytics Tools
Python, R, SQL, Excel VBA, SAS Viya, Power BI, Tableau, Looker Studio
Data and Engineering
PySpark, Docker, Airflow, MLflow, Pandas, NumPy, Git
Machine Learning
XGBoost, LightGBM, Random Forest, Elastic Net, K-means Clustering, Time Series Forecasting, TensorFlow, Scikit-learn
Business Systems
SAP Business One, e-commerce analytics, forecasting, inventory optimization, campaign performance reporting
Machine Learning Workflow
1. Problem Framing
Define the business objective, target variable, decision point, and success metrics before choosing the model.
2. Data Preparation
Clean, join, validate, and document datasets so the modeling base is reproducible and defensible.
3. Feature Engineering
Create predictive signals from transactions, time periods, categories, customer behavior, and operational constraints.
4. Model Development
Train and compare statistical, tree-based, boosting, clustering, and deep learning approaches based on the problem type.
5. Evaluation
Assess model performance with appropriate metrics, error analysis, validation strategy, and business cost trade-offs.
6. Decision Output
Translate results into recommendations, dashboards, slide narratives, or implementation-ready next steps.
Education
Singapore Management University
Master of IT in Business
Data Science and Analytics Track
Aug 2025 - Dec 2026
University of London (SIM)
Bachelor of Science in Business and Management
Second-Upper Class Division
Aug 2014 - Aug 2017
Featured Data Science Projects
Click on the project title or button to view the project documents. Best viewed on a computer.
Demand Forecasting
Inventory Optimization
Demand Forecasting and Inventory Optimization
Topics: Python, Elastic Net, Random Forest, LightGBM, Prophet, SKU Prioritization
Built a demand forecasting and inventory optimization workflow on Walmart POS data, benchmarking Elastic Net, Random Forest, LightGBM, and Prophet with sliding-window validation, recursive 2016 forecasting, and capacity-constrained SKU prioritization.
Multi-Label Image
Classification
Multi-Label Image Classification
Topics: TensorFlow, Deep Learning, Computer Vision, Multi-Label Classification
Built a deep learning workflow for image attribute prediction, with a dedicated performance page for model evaluation, label-level metrics, threshold review, and error analysis.
Transport Data
Architecture
Unified Transport and Ridership Analytics Architecture
Topics: Big Data Architecture, Spark, Airflow, HDFS, Hive, Trino, Superset
Designed a CP-first hybrid analytics architecture for transport operations, covering multi-source ingestion, batch processing, on-premise storage, predictive maintenance, ridership analytics, and operational reporting.
Shiny Visual
Analytics App
Shiny Visual Analytics Application
Topics: R Shiny, Quarto, EDA, Customer Segmentation, Predictive Modeling
Built a fintech customer behavior analytics app that combines exploratory analysis, clustering, and predictive modeling to support customer value analysis and interactive decision-making.
Bank Default
Prediction
Bank Default Prediction
Topics: PySpark, Medallion Architecture, XGBoost, Logistic Regression, Model Monitoring
Designed an end-to-end bank default prediction pipeline with Bronze-Silver-Gold feature stores, chronological validation, Champion-Challenger governance, calibrated inference, PSI/CSI drift monitoring, and a Streamlit control dashboard.
Delivery Delay
Risk Prediction
E-Commerce Delivery Delay Risk Prediction
Topics: Airflow, MLflow, XGBoost, Random Forest, Logistic Regression, Evidently, Streamlit
Operationalized late-delivery risk prediction for Olist e-commerce orders with a medallion data pipeline, calibrated model comparison, Airflow batch inference, Evidently monitoring, alert-driven retraining, and a fulfillment dashboard for logistics and customer service teams.
Project Archive
My full project archive is organized in the navigation bar by tool: R, SAS Viya, and Python. I keep this About page focused on the projects most relevant to data science roles, while the tool pages provide the supporting reports, slides, and application viewers.
Certifications and Languages
Certification: SAP Business One, SAP Certified Application Associate
Languages: English, Mandarin