US 12,293,420 B2
Autonomous accounting anomaly detection engine
Nikil Kumar Reddy, Hyderabad (IN); Swami Subramanian, Houston, TX (US); Sudakshina Chaudhury, Bhubaneswar (IN); and Rohit Roy Burman, Bhubaneswar (IN)
Assigned to HighRadius Corporation, Houston, TX (US)
Filed by HIGHRADIUS CORPORATION, Houston, TX (US)
Filed on Jun. 3, 2022, as Appl. No. 17/832,334.
Claims priority of provisional application 63/294,267, filed on Dec. 28, 2021.
Prior Publication US 2023/0206348 A1, Jun. 29, 2023
Int. Cl. G06Q 40/12 (2023.01); G06Q 20/40 (2012.01)
CPC G06Q 40/128 (2013.12) [G06Q 20/4016 (2013.01); G06Q 20/407 (2013.01); G06Q 40/125 (2013.12)] 19 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, using a computing system, one or more first user inputs into a general ledger stored on the computing system;
analyzing, using the computing system, the one or more first user inputs to detect one or more anomalous transactions associated with the one or more first user inputs, wherein analyzing the one or more first user inputs comprises:
obtaining, using the computing system, historical information associated with one or more previous inputs; and
correlating, using the computing system and one or more machine learning algorithms, the historical information and the one or more first user inputs to detect one or more anomalies associated with the one or more first user inputs;
based on the correlation between the historical information and the one or more first user inputs, detecting, using the computing system, at least one anomalous transaction associated with the one or more first user inputs;
generating, using the computing system, one or more first recommended actions to correct the at least one anomalous transaction;
retraining, using the computer system, the one or more machine learning algorithms based on a subsequent risk score associated with one or more anomalies and one or more corrected user inputs associated with the one or more anomalies;
generating, using the computing system, one or more second recommended actions comprising at least one second recommended action different from the one or more first recommended actions to correct the at least one anomalous transaction using the one or more machine learning algorithms that were retrained;
based on the one or more second recommended actions to correct the at least one anomalous transaction, automatically prioritizing, using the computing system, the one or more second recommended actions; and
based on the prioritized one or more second recommended actions, automatically executing, using the computing system, a prioritized second recommended action having a highest priority to correct the at least one anomalous transaction.