Case Study: Arbetsförmedlingen achieves real-time job recommendations and discriminatory text detection with Hopsworks

Job Recommendations & NLP processing on Job ADs

Arbetsförmedlingen, the Swedish Public Employment Service, needed a highly available production environment to manage AI workflows for job recommendations and NLP processing on job ads. They turned to Hopsworks to support GPU access, training and serving orchestration, and real-time predictions for suitable job postings for about 700,000 new job seekers each year.

Hopsworks provided a scalable model serving platform with online and offline models, data lake integrations, and GDPR-compliant data sharing. With Hopsworks, Arbetsförmedlingen improved centralization and collaboration for data scientists, enabling more structured feature pipelines and AI model development while also helping identify discriminatory text in job announcements.

Arbetsförmedlingen