Abstract: The online concierge system generates task units based on orders and assigns batches of task units to pickers. The online concierge system generates task units based on received orders. The online concierge system generates permutations of these task units to generate candidate sets of task batches. The online concierge system scores each of these candidate sets, and selects a set of task batches to assign to pickers based on the scores. Additionally, to determine which task UI to display to the picker, the picker client device uses a UI state machine. The UI state machine is a state machine where each state corresponds to a task UI to display on the picker client device. The state transitions between the UI states of the UI state machine indicate which UI state to transition to from a current UI state based on the next task unit in the received task batch.

Abstract: An image-based recommendation system provides a first set of selectable images to a user and receives user selections from the first set of images. If a user selects a single image from the first set of images, the product page associated with the selected image is transmitted to the user. If the user makes multiple selections from the first set of images, then the images associated with the user selections are analyzed and a second set of similar images is generated for transmission to the user. The process of receiving and analyzing user-selected images, generating image sets and transmission of generated images sets continues until the user selects a single image. The precision of identification of similar images can be improved by providing the user selections as training data to the image-based recommendation system.

Abstract: An image-based recommendation system provides a first set of selectable images to a user and receives user selections from the first set of images. If a user selects a single image from the first set of images, the product page associated with the selected image is transmitted to the user. If the user makes multiple selections from the first set of images, then the images associated with the user selections are analyzed and a second set of similar images is generated for transmission to the user. The process of receiving and analyzing user-selected images, generating image sets and transmission of generated images sets continues until the user selects a single image. The precision of identification of similar images can be improved by providing the user selections as training data to the image-based recommendation system.