Neural Network–Enhanced Inventory Optimization for Retail
I developed a neural network–based model to enhance traditional Economic Order Quantity (EOQ) inventory management for Tim Hortons China. The objective was to address the challenges of frequent product launches and cost pressures in a highly competitive retail environment. The project involved preprocessing and structuring business data, including sales, pricing, and demand-related variables. I performed data cleaning, categorization, and feature engineering to improve data quality and model performance, which aligns closely with data labeling and annotation workflows. Additionally, I evaluated model outputs, identified inconsistencies, and iteratively refined inputs to improve predictive accuracy and decision-making reliability. This experience is directly relevant to AI training tasks such as data annotation, response evaluation, and quality control. I applied strong analytical reasoning to assess model behavior and ensure consistency between inputs and outputs, demonstrating the ability to provide structured, high-quality feedback for model improvement.