AmazonGranted patentAI & softwareInterest 6/10

Amazon patents system that guesses if you are shopping for work or home

This patent describes a method for online stores to automatically detect if a user is browsing with a personal or business account. It uses machine learning models to analyze the item being viewed and the user's past behavior to predict which account type is appropriate. If the system predicts a business account, it displays a specific template or interface designed for that account type over the current page.

Why it mattersThis could make online shopping smoother for people who use the same website for both personal and professional purchases by automatically adjusting the interface to match their intent.

In the filing’s own words

Techniques are described herein for determining whether the account type to use is the first account type or the second account type. An example method includes receiving a request associated with an item, the online session associated with an account identifier and a first account having a first account type. Generating a prediction as to whether the whether the account type to use is the first account type or second account type based at least in part on whether the item is processed differently using the first account or the second account. Determining, based on the prediction and whether the account identifier is associated with the second account having the second account type, whether to display a first template associated with the second account. Transmitting a message to display the template.

What exactly they claim (claim 1)

1. A method performed by a computing system, the method comprising: receiving, via a graphical user interface (GUI) of a user device and during an online session with the user device, a request associated with an item displayed on the GUI, the online session associated with an account identifier and a first account; generating a first machine learning input associated with the item based at least in part on the request; accessing historical information associated with the account identifier based at least in part on receiving the request; generating a second machine learning input based on the historical information; generating, using a first machine learning model and based on the first machine learning input, a first prediction as to whether an account type to use is a first account type or a second account type, the request to be processed differently using the first account type than the second account type; generating, using a second machine learning model and based on the second machine learning input, a second prediction as to whether the account type to use is the first account type or the second account type; generating, using a third machine learning model and based on the first prediction and the second prediction, a combined prediction that the account type to use is the second account type; determining, based on the combined prediction, a template associated with the second account type to be displayed on the GUI; generating a message to display the template associated with the second account type on the GUI; and causing, via the message, the user device to display the template associated with the second account type on the GUI during the online session, wherein the message includes instructions to overlay the template associated with the second account type onto underlying content associated with the item.

Facts

Company
Amazon (Amazon Technologies, Inc.)
Official title
Real time account type identification
Patent number
US12792974B1
Application number
18474871
Filed
Sep 26, 2023
Granted
Nov 10, 2026
Inventors
2
Claims
20

Classified under

G06N 20/20 Physics & computingG06Q 10/0838 Physics & computing

More from Amazon in AI & software

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.