US 11,842,740 B2
Seamless authentication and enrollment
Shahar Faians, Tel Aviv (IL); Avraham Lousky, Hod Hasharon (IL); Elad Hoffman, Tel Aviv (IL); Alon Moshe Sabban, Rishon le Zion (IL); Jade Tarni Kahn, Givat Shmuel (IL); and Roie Mandler, Kfar-Saba (IL)
Assigned to NICE LTD., Ra'anana (IL)
Filed by NICE LTD., Ra'anana (IL)
Filed on Oct. 15, 2020, as Appl. No. 17/071,777.
Application 17/071,777 is a continuation of application No. 15/436,020, filed on Feb. 17, 2017, granted, now 10,854,204.
Application 15/436,020 is a continuation of application No. 13/875,317, filed on May 2, 2013, granted, now 9,620,123, issued on Apr. 11, 2017.
Prior Publication US 2021/0074301 A1, Mar. 11, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 17/04 (2013.01); G06Q 30/018 (2023.01); G10L 17/06 (2013.01); G10L 17/22 (2013.01); H04L 9/40 (2022.01)
CPC G10L 17/04 (2013.01) [G06Q 30/0185 (2013.01); G10L 17/06 (2013.01); G10L 17/22 (2013.01); H04L 63/0861 (2013.01)] 26 Claims
OG exemplary drawing
 
17. An apparatus comprising:
at least one processor configured to:
receive a plurality of authenticated audio streams from an audio server that archives audio data of historical communication sessions between a customer and a human representative prior to a call by a customer
wherein authenticated audio streams were authenticated based on historical transactional data, historical interaction data, or both;
generate by a processor a voice print for the customer based on the plurality of authenticated audio streams for the customer with the human representative;
automatically, by the processor, authenticate the customer in a subsequent communication while the customer is talking to the human representative using the voice print, wherein the authenticating is performed without requesting authentication details from the customer or interrupting the call;
receive an additional audio stream from the audio server of a second historical communication session between the customer and the human representative;
create a preliminary association between the additional audio stream and the customer;
determine a confidence level of the preliminary association based on stored authentication information of the customer;
if the confidence level is higher than a threshold, checking using the processor the identity of the customer against a suspicious customer list, and if the identity of the customer is not listed in the suspicious customer list, sending a request to compare the audio stream to a database of voice prints of known fraudsters;
performing the operation of sending a request to generate from the audio stream a current voice print associated with the customer only if the audio stream does not match any known fraudsters; and
if the current voice print does not match any other voice prints of other customers and the current voice print is consistent with the customer's voice print, then enroll the current voice print for the customer in a customer voice print database.