Case Study: Western Digital achieves smarter email categorization and quality control with Dataiku

Western Digital Smarter Email Categorization With NLP

Western Digital, a global storage and semiconductor company, faced a high-volume manual email triage process in its logistics control tower, with 8,000–10,000 emails a week slowing response times and making it hard to prioritize critical issues. Using Dataiku’s NLP and MLOps capabilities, the company also aimed to improve wafer-level quality control by detecting defects earlier in production.

Dataiku helped Western Digital build an automated email categorization and analytics workflow that extracts, labels, and analyzes about 10,000 emails per week with over 80% accuracy. The solution cut email traffic by 17%, saved about 100 employee hours per month, and reduced response time by more than 20 hours, while also enabling automated outlier detection in wafer testing to reduce waste, improve yield, and strengthen product quality.

Western Digital