Abstract: Transaction modification based on modeled profiles is described. In an example, transaction data can be received from merchant computing devices associated with merchants associated with a payment processing system. A model can be trained to generate profiles using, as training data, one or more of merchant data, buyer data, or the transaction data. Upon receiving an indication of a particular transaction between a buyer and a merchant, it can be determined that a characteristic of the transaction corresponds to a profile of the generated profiles. Based on the determination that the characteristic corresponds to the profile and the transaction data, a recommendation can be generated for a modification of the transaction to add an item or replace an item.

Abstract: Transaction modification based on modeled profiles is described. In an example, transaction data can be received by a payment processing system. The payment processing system can generate, using a trained model, profiles. Based at least in part on (i) receiving an indication that a buyer and a merchant are engaged in, or likely to be engaged in, a transaction and (ii) determining that a characteristic of the transaction corresponds to a profile, generating a communication to enable the buyer or the merchant to modify the transaction in association with processing of the transaction.

Abstract: Transaction modification based on modeled profiles is described. In an example, transaction data can be received by a payment processing system. The payment processing system can generate, using a trained model, profiles. Based at least in part on (i) receiving an indication that a buyer and a merchant are engaged in, or likely to be engaged in, a transaction and (ii) determining that a characteristic of the transaction corresponds to a profile, generating a communication to enable the buyer or the merchant to modify the transaction in association with processing of the transaction.

Abstract: Buyer profile management is described herein. In an example, a payment processing service can store, in a database, buyer profiles associated with buyers involved in transactions with one or more merchants of the payment processing service. The payment processing service can receive transaction data associated with a new transaction between a buyer and a merchant and can determine whether the transaction data corresponds to any of the buyer profiles. Based at least in part on determining that the transaction data corresponds to one of the buyer profiles, the payment processing service can update the database storing the buyer profiles, wherein updating the database comprises merging two or more buyer profiles. A recommendation can be made based on the merged buyer profile prior to completion of the new transaction.

Abstract: Buyer profile management is described herein. In an example, a payment processing service can store, in a database, buyer profiles associated with buyers involved in point-of-sale (POS) transactions with one or more merchants of the payment processing service. The payment processing service can receive transaction data associated with a new POS transaction between a buyer and a merchant and can determine whether the transaction data corresponds to any of the buyer profiles. Based at least in part on determining whether the transaction data corresponds to any of the buyer profiles, the payment processing service can update the database storing the buyer profiles, wherein updating the database comprises (i) associating the transaction data with an existing buyer profile, (ii) merging two or more buyer profiles, or (iii) adding a new buyer profile.

Abstract: Techniques for providing items recommendations based on a state of a merchant device are described. For example, a determination of whether or not a merchant device is in an online mode or an offline mode may be made. When it is determined that the merchant device is in the online mode, a recommendation that is received for the current transaction may be presented. When it is determined that the merchant device is in the offline mode, a recommendation that is previously received may be presented. In other examples, other types of recommendations may be presented.

Abstract: Techniques for providing items recommendations based on a state of a merchant device are described. For example, a determination of whether or not a merchant device is in an online mode or an offline mode may be made. When it is determined that the merchant device is in the online mode, a recommendation that is received for the current transaction may be presented. When it is determined that the merchant device is in the offline mode, a recommendation that is previously received may be presented. In other examples, other types of recommendations may be presented.

Abstract: A service provider may receive transaction information from a plurality of merchant devices for a plurality of point of sale (POS) transactions conducted between buyers and merchants. The service provider associates the transaction information with respective buyer profiles. The service provider may subsequently receive an indication of a first item selected by a particular buyer for a current POS transaction being conducted with a particular merchant. The service provider may determine a particular buyer profile to associate with the current transaction. Based at least in part on the particular buyer profile or the other buyer profiles, the service provider may determine a second item that has been purchased frequently together with the first item. The service provider sends a communication to the particular merchant to indicate that the particular merchant may wish to offer the second item to the particular buyer for inclusion in the current transaction.