Singapore, Singapore
Assisted undergraduate students in answering their doubts regarding mathematical concepts as well as guiding them in their courseworks. Related courseworks: MH1101 Calculus II MH1201 Linear Algebra II MH1301 Discrete Mathematics MH1403 Algorithms and Computing MH3110 Ordinary Differential Equations MH3500 Statistics MH3511 Data Analysis with Computer PS0002 Introduction to Data Science and Artificial Intelligence SC1005 Digital Logic SC1006 Computer Organisation and Architecture SC2005 Operating Systems SC2006 Software Engineering SC2207 Database Systems Principles
• Developed an optimized architecture for large-scale data cleaning and transformation, integrating over 1,500 CSV files with inconsistent header naming into a master dataset with standardized headers. Leveraged Microsoft Fabric Notebook, Dataflow and Datalake, achieving a 4000% reduction in processing time compared to the previous design. Incorporated Microsoft Graph API and Azure Active Directory tokens to enable seamless integration between SharePoint and Fabric Notebook • Deployed a Power BI dashboard by integrating advanced statistical techniques including ANOVA, Tukey’s HSD, Wilcoxon Signed-Rank test, Friedman test, and Nemenyi post-hoc test to evaluate if taste profiles of two beverages were statistically similar using hypothesis testing. Achieved 99.7% accuracy when benchmarked against XLSTAT Macro files • Enhanced an automated web crawling pipeline using Selenium and ChromeDriver to scrape F&B outlet data across Singapore’s planning areas. Integrated NLP and fuzzy matching techniques to map scraped business categories to existing categories in F&N’s database, improving efficiency by 20% • Utilized OpenAI GPT API to automate PDF-to-text extraction, standardize dataset formatting. Consolidated raw data into a structured master table with consistent brand columns, enabling analysis of emerging beverages for trend prediction. Identified and merged 15% redundant brand entries into larger parent groups, reducing redundant naming • Improved an object detection project using Microsoft Custom Vision by performing necessary data augmentation and manually tagging over 300 images to train a computer vision model for detecting beverages in F&N chillers. Improved accuracy by 40% compared to previous model • Collaborated with business stakeholders to gather requirements through meetings and developed impactful Power BI dashboards and automation pipelines. Delivered production-ready dashboards that supported informed decision-making and streamlined business operations