- Project 1: Researched and developed a CNN-based wafer pattern recognition model that can handle data from different sources with different numbers of inspection points. - Project 2: Enhanced their current ML model effectively by addressing performance issues when transferring the model to new product predictions. - Project 3: Engaged in full-stack development of the company's internal reporting system through Azure DevOps.

Developed an investment report analyzer that effectively summarizes, tags, and analyzes investment reports for investment analysts.

Enhanced the ML defect prediction model by significantly increasing the forecast F1 score from the current 20% to 70%.