Granted Patent B2

US 12,493,619 · App. 18/748,448 · Granted Dec 9, 2025

Inventors: Nishant Srivastava (Glen Allen, VA); Joshua Condon (Maidens, VA); Eric K. Barnum (Midlothian, VA)

Assignee: Capital One Services, LLC

G06F16/24568G06F9/467

Abstract

Claims (61)

1 . A computer-implemented method comprising:

receiving, by a streaming data platform, a transaction object corresponding to a transaction, wherein the transaction object comprises transaction metadata indicating a workflow corresponding to a transaction type of the transaction object, wherein the workflow comprises a first plurality of processing steps to process the transaction;

determining, by a classification microservice and based on transaction details associated with the transaction object, whether the transaction object comprises information that would allow the transaction object to be processed via an alternative workflow, wherein the alternative workflow comprises a second plurality of processing steps to process the transaction differently from the first plurality of processing steps;

changing, by the classification microservice, the indication of the workflow to the alternative workflow;

processing, by a microservice associated with the alternative workflow and based on a determination that a current workflow stage of the transaction object matches a first workflow stage associated with the microservice, the transaction object;

determining that the current workflow stage of the transaction object indicates that the transaction object has completed processing corresponding to the alternative workflow; and

removing the transaction object from the streaming data platform and outputting the transaction object and an indication that the transaction object has completed the processing corresponding to the alternative workflow to a downstream system.

2 . The computer-implemented method of claim 1 , wherein the classification microservice uses one or more machine-learning models when determining whether the transaction object can be processed via an alternative workflow.

3 . The computer-implemented method of claim 2 , further comprising:

training the one or more machine-learning models based on a plurality of previously processed transaction objects, wherein each of the plurality of previously processed transaction objects comprises a transaction score.

4 . The computer-implemented method of claim 2 , further comprising:

dividing a plurality of previously processed transaction objects into a plurality of transaction clusters corresponding to different transaction types; and

training the one or more machine-learning models using different sets of the plurality of transaction clusters.

5 . The computer-implemented method of claim 1 , further comprising:

determining, by the classification microservice and using one or more machine-learning models, a first transaction score for processing the transaction object according to the workflow corresponding to the transaction type; and

determining, by the classification microservice and using the one or more machine-learning models, a second transaction score for processing the transaction object according to the alternative workflow corresponding to the transaction type, wherein the determination that the transaction object can be processed via an alternative workflow is further based on an indication that the second transaction score represents a more cost-effective approach than the first transaction score.

6 . The computer-implemented method of claim 1 , wherein the determination that the transaction object can be processed via an alternative workflow is further based on at least one of: a payment value of the transaction, a deadline for the transaction, a security level of the transaction, or a cost of the transaction.

7 . The computer-implemented method of claim 1 , wherein the determination that the transaction object can be processed via an alternative workflow further comprises at least one of:

determining a transaction deadline for the transaction; or

determining, using one or more machine-learning models, an estimated time of completion for the transaction.

8 . One or more non-transitory computer readable media comprising instructions that, when executed by at least one processor, cause a computing device to perform operations comprising:

receiving a transaction object corresponding to a transaction, wherein the transaction object comprises transaction metadata indicating a workflow corresponding to a transaction type of the transaction object, wherein the workflow comprises a first plurality of processing steps to process the transaction;

determining, by a classification microservice and based on transaction details associated with the transaction object, whether the transaction object comprises information that would allow the transaction object to be processed via an alternative workflow, wherein the alternative workflow comprises a second plurality of processing steps to process the transaction differently from the first plurality of processing steps;

changing, by the classification microservice, the indication of the workflow to the alternative workflow;

processing, by a microservice associated with the alternative workflow and based on a determination that a current workflow stage of the transaction object matches a first workflow stage associated with the microservice, the transaction object;

determining that the current workflow stage of the transaction object indicates that the transaction object has completed processing corresponding to the alternative workflow; and

outputting the transaction object with an indication that the transaction object has completed the processing corresponding to the alternative workflow to a downstream system.

9 . The one or more non-transitory computer readable media of claim 8 , wherein the classification microservice uses one or more machine-learning models when determining whether the transaction object can be processed via an alternative workflow.

10 . The one or more non-transitory computer readable media of claim 9 , wherein the instructions, when executed by the at least one processor, cause the computing device to perform operations further comprising:

training the one or more machine-learning models based on a plurality of previously processed transaction objects, wherein each of the plurality of previously processed transaction objects comprises a transaction score.

11 . The one or more non-transitory computer readable media of claim 9 , wherein the instructions, when executed by the at least one processor, cause the computing device to perform operations further comprising:

dividing a plurality of previously processed transaction objects into a plurality of transaction clusters corresponding to different transaction types; and

training the one or more machine-learning models using different sets of the plurality of transaction clusters.

12 . The one or more non-transitory computer readable media of claim 9 , wherein the determination that the transaction object can be processed via an alternative workflow is further based on a payment value of the transaction, a deadline for the transaction, a security level of the transaction, or a cost of the transaction.

13 . The one or more non-transitory computer readable media of claim 8 , wherein the instructions, when executed by the at least one processor, cause the computing device to perform operations further comprising:

determining, by the classification microservice and using one or more machine-learning models, a first transaction score for processing the transaction object according to the workflow corresponding to the transaction type; and

determining, by the classification microservice and using the one or more machine-learning models, a second transaction score for processing the transaction object according to the alternative workflow corresponding to the transaction type, wherein the determination that the transaction object can be processed via an alternative workflow is further based on an indication that the second transaction score represents a more cost-effective approach than the first transaction score.

14 . The one or more non-transitory computer readable media of claim 8 , wherein the instructions, when executed by the at least one processor, cause the computing device to determine that the transaction object can be processed via an alternative workflow based on at least one of:

determining a transaction deadline for the transaction; or

determining, using one or more machine-learning models, an estimated time of completion for the transaction.

15 . A transaction exchange platform comprising:

a streaming data platform;

a plurality of microservices comprising at least a classification microservice and a microservice, wherein each microservice of the plurality of microservices is configured to watch for transactions on the streaming data platform in a corresponding workflow stage based on a plurality of workflows corresponding to a plurality of transaction types;

at least one processor; and

memory storing instructions that, when executed by the at least one processor, cause the transaction exchange platform to:

receive, by the streaming data platform, a transaction object corresponding to a transaction, wherein the transaction object comprises transaction metadata indicating a workflow corresponding to a transaction type of the transaction object, wherein the workflow comprises a first plurality of processing steps to process the transaction;

determine, by the classification microservice and based on transaction details associated with the transaction object, whether the transaction object comprises information that would allow the transaction object to be processed via an alternative workflow, wherein the alternative workflow comprises a second plurality of processing steps to process the transaction differently from the first plurality of processing steps;

change, by the classification microservice, the indication of the workflow to the alternative workflow;

process, by a microservice associated with the alternative workflow and based on a determination that a current workflow stage of the transaction object matches a first workflow stage associated with the microservice, the transaction object;

determine that the current workflow stage of the transaction object indicates that the transaction object has completed processing corresponding to the alternative workflow; and

remove the transaction object from the streaming data platform and output the transaction object and an indication that the transaction object has completed the processing corresponding to the alternative workflow to a downstream system.

16 . The transaction exchange platform of claim 15 , wherein the classification microservice uses one or more machine-learning models when determining whether the transaction object can be processed via an alternative workflow.

17 . The transaction exchange platform of claim 16 , wherein the instructions, when executed by the at least one processor, further cause the transaction exchange platform to:

train the one or more machine-learning models based on a plurality of previously processed transaction objects, wherein each of the plurality of previously processed transaction objects comprises a transaction score.

18 . The transaction exchange platform of claim 16 , wherein the instructions, when executed by the at least one processor, further cause the transaction exchange platform to:

divide a plurality of previously processed transaction objects into a plurality of transaction clusters corresponding to different transaction types; and

train the one or more machine-learning models using different sets of the plurality of transaction clusters.

19 . The transaction exchange platform of claim 15 , wherein the instructions, when executed by the at least one processor, further cause the transaction exchange platform to:

determine, using one or more machine-learning models, a first transaction score for processing the transaction object according to the workflow corresponding to the transaction type; and

determine, using the one or more machine-learning models, a second transaction score for processing the transaction object according to the alternative workflow corresponding to the transaction type, wherein the determination that the transaction object can be processed via an alternative workflow is further based on an indication that the second transaction score represents a more cost-effective approach than the first transaction score.

20 . The transaction exchange platform of claim 15 , wherein the determination that the transaction object can be processed via an alternative workflow is further based on at least one of a payment value of the transaction, a deadline for the transaction, a security level of the transaction, or a cost of the transaction.

Assignments (1)

Continuity (2)

References Cited (62)