Maplebear Inc.

Shishir Kumar Prasad

Sharath Rao Karikurve

Prithvishankar Srinivasan

Shih-Ting Lin

Sharath Rao

Diego Goyret

Haixun Wang

Tejaswi Tenneti

Min Xie

Tyler Russell Tate

Saurav Manchanda

Jonathan Lennart Bender

Joey Loi

Jeremy Stanley

Qingyuan Chen

Riddhima Sejpal

Abhay Pawar

Cameron Nicholas Taylor

John Salaveria

Negin Entezari

Luis Manrique

Kevin McCullough

Embodiments relate to utilizing a language model to automatically generate a novel recipe with refined content, which can be offered to a user of an online system. The online system generates a first prompt for input into a large language model (LLM), the first prompt including a plurality of task requests for generating initial content of a recipe. The online system requests the LLM to generate, based on the first prompt input into the LLM, the initial content of the recipe. The online system generates a second prompt for input into the LLM, the second prompt including the initial content of the recipe and contextual information about the recipe. The online system requests the LLM to generate, based on the second prompt input into the LLM, refined content of the recipe. The online system stores the recipe with the refined content in a database of the online system.

Using language model to generate recipe with refined content

An online concierge system generates suggested instructions for presentation to a user. The online concierge system access instruction examples corresponding to a target item category and generates candidate instruction representations based on instruction messages within each instruction example. The online concierge system generates preliminary scores for the candidate instruction representations that are directly related to an intra-category frequency of use of the instruction tokens of the candidate instruction representation within the target item category. The online system normalizes these preliminary scores for the candidate instruction representations based on the inter-category frequency of use of the instruction tokens in all item categories to generate final scores for the candidate instruction representations. The online concierge system selects a set of instruction representations based on these final scores and generates suggested instructions based on the set of instruction representations.

Generating suggested instructions through natural language processing of instruction examples

A computer system uses clustering and a large language model (LLM) to normalize attribute tuples for items stored in a database of an online system. The online system collects attribute tuples, each attribute tuple comprising an attribute type and an attribute value for an item. The online system initially clusters the attribute tuples into a first plurality of clusters. The online system generates prompts for input into the LLM, each prompt including a subset of attribute tuples grouped into a respective cluster of the first plurality. Based on the prompts, the LLM generates a second plurality of clusters, each cluster including one or more attribute tuples that have a common attribute type and a common attribute value. The online system maps each attribute tuple to a respective normalized attribute tuple associated with each cluster. The online system rewrites each attribute tuple in the database to a corresponding normalized attribute tuple.

Using unsupervised clustering and language model to normalize attribute tuples of items in a database

An online concierge system suggests subsequent search queries based on previous search queries and whether the previous search queries resulted in conversions. The online concierge system trains a machine learning model using previous delivery orders and whether initial and subsequent search queries in the previous delivery orders resulted in conversions. When the online concierge system receives a search query to identify one or more items from a customer, the online concierge system parses the search query into combinations of terms and identifies items related to the search query. In response to the search query resulting in a conversion, the online concierge system retrieves a conversion graph and presents a suggested subsequent search query based on the conversion graph. In response to the search query not resulting in a conversion, the online concierge system retrieves a non-conversion graph and presents a suggested subsequent search query based on the non-conversion graph.

Providing search suggestions based on previous searches and conversions

An online concierge system may determine recommended search terms for a user. The online concierge system may receive a request from a user to view a user interface configured to receive a search query. The online concierge system retrieves long-term activity data including previous search terms entered by the user while searching for items to add to an online shopping cart. For each previous search term, the online concierge system retrieves categorical search terms corresponding to one or more categories to which the previous search term was mapped. The online concierge system determines a set of nearby categorical search terms and sends, for display via a client device, the set of nearby categorical search terms as recommended search terms.

Method, non-transitory computer-readable medium, and system for determining recommended search terms for a user of an online concierge system

In an online concierge system, a shopper retrieves items specified in an order by a customer from a retail location. The online concierge system optimizes order fulfillment by selecting a retail location for an order that is most time-efficient and that is most likely to have each of the item in the order available. Hence, the online concierge system may select a less convenient retail location that is more likely to have each item being ordered available. To predict whether a retail location incompletely fulfill the order if selected to fulfill the order, the online concierge system trains a machine learning model based on prior orders fulfilled by the retail location, a shopper retrieving items in the order, items in the order, and other features.

Selecting a location for order fulfillment based on machine learning model prediction of incomplete fulfillment of the order for different locations

An online concierge system allows users to order items from a warehouse having multiple physical locations, allowing a user to order items at any given warehouse location. To select a warehouse location for a warehouse selected by a user, the online concierge system identifies a set of items that the user has a threshold likelihood of purchasing from prior orders by the user. For each of a set of warehouse locations, the online concierge system applies a machine-learned item availability model to each item of the identified set. From the availabilities of items of the set at each warehouse location of the set, the online concierge system selects a warehouse location. The online concierge system identifies an inventory of items from the selected warehouse location to the user for inclusion in an order.

Online shopping system and method for selecting a warehouse for inventory based on predicted availability and predicted replacement machine learning models

An online concierge shopping system identifies candidate items to a user for inclusion in an order based on prior user inclusion of items in orders and items currently included in the order. From a multi-dimensional tensor generated from cooccurrences of items in orders from various users, the online concierge system generates item embeddings and user embeddings in a common latent space by decomposing the multi-dimensional tensor. From items included in an order, the online concierge system generates an order embedding from item embeddings of the items included in the order. Scores for candidate items are determined based on similarity of item embeddings for the candidate items to the order embedding. Candidate items are selected based on their scores, with the selected candidate items identified to the user.

Personalized recommendation of complementary items to a user for inclusion in an order for fulfillment by an online concierge system based on embeddings for a user and for items

A method for predicting inventory availability, involving receiving a delivery order including a plurality of items and a delivery location, and identifying a warehouse for picking the plurality of items. The method retrieves a machine-learned model that predicts a probability that an item is available at the warehouse. The machine-learned model is trained, using machine learning, based in part on a plurality of datasets. The plurality of datasets include data describing items included in previous delivery orders, whether each item in each previous delivery order was picked, a warehouse associated with each previous delivery order, and a plurality of characteristics associated with each of the items. The method predicts the probability that one of the plurality of items in the delivery order is available at the warehouse, and generates an instruction to a picker based on the probability. An instruction is transmitted to a mobile device of the picker.

Growing community of inventors

Fremont, CA, United States of America

Shishir Kumar Prasad