Nordstrom deployed an internal machine learning system called The Outfitter that automatically assembles complete outfit recommendations starting from a single 'seed' product, helping digital stylists scale their work across Nordstrom's catalog.

Built and deployed by Nordstrom's own engineering team, The Outfitter uses a two-step deep learning process: first it generates an outfit template of compatible clothing categories based on historical stylist-created outfits, then it selects specific items to fill each slot. According to a Nordstrom engineering blog post, the tool had published hundreds of stylist-quality outfits and covered 87% of Nordstrom's online product catalog. The system was designed to help digital stylists handle peak seasonal demand without reducing quality.