Stephen Chen

Developed and trained a PyTorch CNN to classify 10 automotive part categories, achieving 89.5% test accuracy. Built an end-to-end computer vision pipeline for YOLO-to-classification preprocessing, model training, checkpointing, and evaluation using classification reports and confusion matrices. Implemented image inference to generate top-3 class predictions and confidence scores for user-uploaded automotive part images.

ML Development Intern at IDT, EIT (2026-05 – 2026-08)

• Developed and trained a PyTorch CNN to classify 10 automotive part categories, achieving 89.5% test accuracy.
• Built an end-to-end computer vision pipeline for YOLO-to-classification preprocessing, model training, checkpointing, and evaluation using classification reports and confusion matrices.
• Implemented image inference to generate top-3 class predictions and confidence scores for user-uploaded automotive part images.

Marketing Executive at Neo Developer League (2024-08 – 2024-11)

• Generated 300+ registrations across 10 Waterloo schools through targeted outreach and school partnerships.
• Coordinated kickoff events, check-ins, workshops, and demo day to keep participating teams on schedule.

Undergraduate in Honours Mathematics – University of Waterloo