At Coupang CPLB, I developed a Python- and OCR-based process that validates invoices, packing lists, and certificates of origin, reducing manual processing time per shipment from approximately 30–60 minutes to 30 seconds. I also analyze demand, stockout-related sales losses, overstock risk, and vendor performance to support inventory planning.
At DMETA, I led AI engineers and data scientists across client-facing projects, from defining business problems through implementation. I led and contributed hands-on to a YOLO-based intelligent CCTV solution and oversaw KISA certification testing, which resulted in certification.
Across roles at Kiss Products, WAI, GP Bio, and others, I developed forecasting, customer segmentation, purchase propensity, and anomaly detection models. My background also includes actuarial and risk analysis at Chubb Korea and KPMG Actuarial Services, alongside hands-on work with Python, SQL, and data platforms.

