Amazon One enhances customer experience with new mobile app

Amazon One streamlines user onboarding through a new mobile application. Customers can now enroll for the palm-recognition service by simply capturing a digital image of their palm within the app. This innovation fosters a seamless in-store experience, particularly for first-time users.

Amazon One simplifies user enrollment for its palm-recognition based identity and payment system. Previously, initial sign-up necessitated a visit to a designated physical location. The introduction of the Amazon One app, downloadable from both the Apple App Store and Google Play Store, empowers users to enroll remotely, eliminating in-store delays for first-time users. This innovation streamlines the checkout process for both customers and retailers, resulting in a more efficient and frictionless in-store experience.

The new Amazon One app facilitates a streamlined user onboarding process. Customers can establish their online profile entirely within the app by leveraging their existing Amazon account credentials. The process involves capturing a digital image of their palm(s) and associating a preferred payment method. Once enrolled, users can leverage Amazon One for various functionalities, including contactless payments, access control, age verification, and loyalty program integration. This functionality extends across a growing network encompassing over 500 Whole Foods Market locations in the U.S., select Amazon stores, and more than 150 third-party establishments including stadiums, airports, fitness centers, and convenience stores. The widespread adoption of Amazon One is evident with over 8 million recorded uses and a user retention rate exceeding 80% at Amazon and Whole Foods Market stores.

The development of Amazon One benefitted from generative AI technology. This technique involved the creation of synthetic palm images, which proved instrumental in training the system's machine learning models. Amazon's commitment to innovation extends to the enrollment process itself. By leveraging the power of artificial intelligence, the system achieves a high degree of accuracy and security. This is accomplished by seamlessly matching a user's camera phone photo with the near-infrared imagery captured by an Amazon One device. This two-factor verification ensures secure enrollment completion only after a successful initial comparison between the user's palm and vein imagery.

Amazon One employs a multi-factor approach to user identification. During enrollment, a unique numerical representation, or 'palm signature,' is generated by analyzing both the user's palm and its underlying vein structure. To uphold the exceptional accuracy of 99.9999%, a novel AI innovation compares vector representations of palm images captured via the Amazon One app with the corresponding vector representation derived from palm and vein images captured by an Amazon One device. This two-step verification process ensures that the individual hovering their palm over the device is the same person who enrolled using the app.

Amazon One prioritizes user privacy and data security. Palm images captured during enrollment via the Amazon One app are encrypted in transit to a secure Amazon One domain within the AWS cloud. These images are not transferable and cannot be saved to user devices. The mobile app incorporates additional layers of spoof detection to further enhance security.