# US 12,111,846 B2 · Computer-based systems configured for machine learning assisted data replication and methods of use thereof > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,111,846 B2 - Document ID: US12111846B2 - Status: Granted - Kind: B2 - Application number: 17/527,626 - Filed: 2021-11-16 - Published: 2023-05-18 - Granted: 2024-10-08 - Earliest claimed: 2021-11-16 - Assignees: Capital One Services, LLC - Applicants: Capital One Services, LLC - Inventors: Ebrima N. Ceesay; Hrishikesh Mukundan Menon; Mohamed Seck - Web page: https://patentgrep.com/patent/US12111846B2 ## CPC - G06F 16/275: {Synchronous replication} - G06F 21/6218: {to a system of files or objects, e.g. local or distributed file system or database} - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06N 20/00: Machine learning - G06N 20/10: using kernel methods, e.g. support vector machines [SVM] - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Systems and methods of data replication via machine learning techniques are disclosed. In one embodiment, an exemplary computer-implemented method may comprise: utilizing a trained replication machine learning model to identify an existing object in the bucket for replication, and a commencing time to replicate the existing object, the commencing time determined based on replication failure predicted by the replication machine learning model; capturing, in response to identifying the existing object for replication, a snapshot of the bucket, the snapshot comprising information related to at least one of: the existing object, metadata of the existing object, and/or an access control list (ACL) of the existing object; and replicating the existing object to a destination cloud according to the determined commencing time, the destination cloud being hosted at a cross-region storage. ## Claims ### Claim 1 A method comprising: obtaining, by one or more processors, a trained replication machine learning model that is trained to: detect at least one object in a bucket at a source cloud for replication, analyze historical replication failure data associated with a plurality of replication failures, and determine a timing for replicating the at least one object; utilizing, by the one or more processors and in response to the bucket being configured with a cross region replication (CRR) service, a replication tool, comprising the trained replication machine learning model, to: identify at least one existing object in the bucket for replication, predict a replication failure based on the historical replication failure data, determine a commencing time to replicate the at least one existing object based on the replication failure, capture at least one snapshot of the bucket based on an identification of the at least one existing object, the at least one snapshot comprising information related to at least one of: the at least one existing object, metadata of the at least one existing object, or an access control list (ACL) of the at least one existing object, generate at least one new object based on the snapshot of the bucket and the replication failure, and replicate, when the replication failure occurs, the at least one new object-to a destination cloud according to the determined commencing time, the destination cloud being hosted at a cross-region storage. ### Claim 2 (depends on 1) The method of claim 1, wherein the replicate the at least one existing object comprises: replicating, by the one or more processors, at least one of: data associated with the at least one existing object, one or more tags associated with the at least one existing object, one or more ACLs associated with the at least one existing object, and one or more encryption states associated with the at least one existing object. ### Claim 3 (depends on 1) The method of claim 1, further comprising: performing, by the one or more processors, integrity check on the replicating the at least one existing object by validating one or more entity tags (e-tags) associated with the at least one existing object. ### Claim 4 (depends on 1) The method of claim 1, wherein the destination cloud is the same cloud utilized in the CRR service. ### Claim 5 (depends on 1) The method of claim 1, further comprising: throttling, by the one or more processors, a replication of the at least one existing object to the destination cloud. ### Claim 6 (depends on 5) The method of claim 5, wherein the throttling is based at least in part on replication failures predicted by the trained replication machine learning model. ### Claim 7 (depends on 1) The method of claim 1, further comprising: providing, by the one or more processors, a graphical user interface for displaying a visualization of a progress associated with the replication the at least one existing object. ### Claim 8 (depends on 1) The method of claim 1, further comprising: generating, by the one or more processors, three copies of data replicated for the at least one existing object. ### Claim 9 (depends on 1) The method of claim 1, wherein the identify the at least one existing object in a bucket comprises: profiling the bucket based on sensitive data; and identifying the at least one existing object based at least in part on the sensitive data. ### Claim 10 (depends on 1) The method of claim 1, wherein the trained replication machine learning model comprises at least one of: a regression model, a classification model, and a recommendation engine model. ### Claim 11 (depends on 1) The method of claim 1, further comprising: determining, by the one or more processors, a critical level for the at least one existing object based at least in part on a classification of sensitive data included in the at least one existing objects; and utilizing, by the one or more processors, the trained replication machine learning model based on the critical level. ### Claim 12 A system comprising: one or more processors; and a memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, cause the one or more processors to: obtain a trained replication machine learning model that is trained to; detect at least one object in a bucket at a source cloud for replication, analyze historical replication failure data associated with a plurality of replication failures, and determine a timing for replicating the at least one object; and a replication tool utilizing the trained replication machine learning model to: identify at least one existing object in the bucket for replication, predict a replication failure based on the historical data, determine a commencing time to replicate the at least one existing object based on the replication failure, capture at least one snapshot of the bucket based on the identification of the at least one existing object, the at least one snapshot comprising information related to at least one of: the at least one existing object, metadata of the at least one existing object, or an access control list (ACL) of the at least one existing object, generate at least one new object based on the at least one snapshot of the bucket and the replication failure, and replicate, when the replication failure occurs, the at least one new object to a destination cloud according to the determined commencing time, the destination cloud being hosted at a cross-region storage. ### Claim 13 (depends on 12) The system of claim 12, wherein to replicate the at least one existing object comprises: replicate at least one of: data associated with the at least one existing object, one or more tags associated with the at least one existing object, one or more ACLs associated with the at least one existing object, and one or more encryption states associated with the at least one existing object. ### Claim 14 (depends on 12) The system of claim 12, wherein the instructions further cause the one or more processors to: perform integrity check on replication of the at least one existing object by validating one or more entity tags (e-tags) associated with the at least one existing object. ### Claim 15 (depends on 12) The system of claim 12, wherein the destination cloud is the same cloud utilized in the CRR service. ### Claim 16 (depends on 12) The system of claim 12, wherein the instructions further cause the one or more processors to: throttle a replication of the existing object to the destination cloud. ### Claim 17 (depends on 16) The system of claim 16, wherein to throttle the replication of the at least one existing objects is based at least in part on replication failures predicted by the trained replication machine learning model. ### Claim 18 (depends on 12) The system of claim 12, wherein the instructions further cause the one or more processors to: profile the bucket based on sensitive data; and identify the at least one existing object based at least in part on the sensitive data. ### Claim 19 (depends on 12) The system of claim 12, wherein the trained replication machine learning model comprises at least one of: a regression model, a classification model, and a recommendation engine model. ### Claim 20 A non-transitory computer readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining the steps of: obtaining, by one or more processors, a trained replication machine learning model that is trained to; detect at least one object in a bucket at a source cloud for replication, analyze historical data associated with replication failure, and a timing for replicating the at least one object; utilizing, by the one or more processors and in response to the bucket being configured with a cross region replication (CRR) service, a replication tool, comprising the trained replication machine learning model, to: identify at least one existing object in the bucket for replication, predict a replication failure based on the historical data, determine a commencing time to replicate the at least one existing object based on the replication failure, capture at least one snapshot of the bucket based on the identification of the at least one existing object, the at least one snapshot comprising information related to at least one of: the at least one existing object, metadata of the at least one existing object, or an access control list (ACL) of the at least one existing object, generate at least one new object based on the snapshot of the bucket and the replication failure, and replicate, when the replication failure occurs, the at least one new object to a destination cloud according to the determined commencing time, the destination cloud being hosted at a cross-region storage.