Search

Maplebear Inc. Granted Patent on Content Eligibility Using Machine Learning

The USPTO granted Patent US12608725B2 to Maplebear Inc. on April 21, 2026, covering an online system that uses machine-learning models to predict user metrics and determine content-item presentation eligibility. The system generates optimal values for users based on predicted metrics, stated objectives, and constraints, then sends content to client devices only when a user qualifies. The patent contains 18 claims and lists Tilman Drerup, Levi Boxell, and Rishikesh Yardi as inventors.