US 11,836,793 B2
Invoice deferral recommendation
Shrihari Vasudevan, Chennai (IN); Sudhanshu Shekhar Singh, New Delhi (IN); Rakesh Rameshrao Pimplikar, Bangalore (IN); Gyana Ranjan Parija, Gurgaon (IN); Jasmina Mohorn, Durham, NC (US); Didier Denove, Berlin (DE); Magesh A Narayanan, New Albany, OH (US); and Khalid Siddiqui, Cracow (PL)
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 11, 2020, as Appl. No. 17/119,026.
Prior Publication US 2022/0188920 A1, Jun. 16, 2022
Int. Cl. G06Q 40/02 (2023.01); G06Q 40/00 (2023.01); G06Q 30/00 (2023.01); G06Q 40/03 (2023.01); G06N 5/043 (2023.01); G06Q 40/12 (2023.01); G06Q 30/016 (2023.01); G06Q 30/04 (2012.01)
CPC G06Q 40/03 (2023.01) [G06N 5/043 (2013.01); G06Q 30/016 (2013.01); G06Q 30/04 (2013.01); G06Q 40/12 (2013.12)] 16 Claims
OG exemplary drawing
 
1. A computer implemented method, comprising:
receiving, at a credit management agent that communicates with a human agent, information corresponding to a customer of a seller, wherein the information is related to credit information of the customer;
generating, at the credit management agent, a credit attribute for the customer with respect to the seller, wherein the generating comprises utilizing a plurality of artificial intelligence agents that each analyze at least a subset of the information to each generate an agent version of the credit attribute, wherein each of the plurality of artificial intelligence agents analyzes a different source of information and analyzes the subset of the information corresponding to the different source of information assigned to the artificial intelligence agent, wherein the generating comprises the plurality of artificial intelligence agents communicating among themselves and generating, via a consensus among the plurality of artificial intelligence agents, the credit attribute by sharing and from the agent version of the credit attribute generated by each of the plurality of artificial intelligence agents, wherein, when communicating among themselves, each of the plurality of artificial intelligence agents keep techniques used to generate the agent version of the credit attribute private from the other of the plurality of artificial intelligence agents by adding noise to the agent version of the credit attribute that is shared among the plurality of artificial intelligence agents to obfuscate the technique for calculating the agent version of the credit attribute; and
recommending, by the credit management agent and to the human agent, a deferral of at least a portion of a pending invoice of the seller for the customer, wherein a value of the deferral is based upon the credit attribute, wherein the value of the deferral is further based upon an allowed amount of deferral over a time period for the human agent, wherein the recommending occurs before the pending invoice is provided to the seller.