# US 2026/0037666 A1 · System and methods for scanning and profiling data for security compliance

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0037666 A1
- Document ID: US20260037666A1
- Status: Published application
- Kind: A1
- Application number: 19/353,323
- Filed: 2025-10-08
- Published: 2026-02-05
- Earliest claimed: 2023-08-01
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Rohit Joshi
- Web page: https://patentgrep.com/patent/US20260037666A1

## CPC

- G06F 21/6245: {Protecting personal data, e.g. for financial or medical purposes}
- G06N 20/00: Machine learning
- H04L 63/0209: {Architectural arrangements, e.g. perimeter networks or demilitarized zones}
- H04L 63/105: {Multiple levels of security}
- H04L 63/205: {involving negotiation or determination of the one or more network security mechanisms to be used, e.g. by negotiation between the client and the server or between peers or by selection according to the capabilities of the entities involved (negotiation of communication capabilities H04L69/24)}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Techniques are disclosed for performing scanning and profiling of stored data to identify the location of data entries that comprise sensitive data that require storage in accordance with enhanced security requirements. The techniques advantageously move a portion of the scanning and profiling operations closer to the data storage locations, and involve deploying processing resources near the locations at which the data is stored rather than relying solely upon centralizing processing at a separate location. This approach has the advantages of reducing network traffic and latency, which can help improve performance and reduce costs, and may also facilitate organizations to better take advantage of distributed computing architectures, which can scale more effectively than centralized processing.

## Claims

### Claim 1

A client device, coupled to a plurality of data storage components that form part of a data network, comprising:
a memory configured to store instructions; and
processing circuitry configured to execute the instructions stored in the memory to cause the client device to:
locally instantiate an application that is configured to enable the client device to access data entries from the plurality of data storage components;
obtaining, via the locally-instantiated application, a location of a set of data entries stored in a data storage component from among the plurality of data storage components;
process, via the locally-instantiated application using first security policy processing operations that are computationally offloaded from a computing device in the data network, a subset of the set of data entries to compute a confidence score associated with the subset of the set of data entries comprising sensitive data;
based on a determination that the confidence score is within a range of confidence score thresholds:
transmit, via the locally-instantiated application, at least a portion of the subset of the set of data entries to the computing device as encrypted and tokenized data,
wherein transmitting the at least the portion of the subset of the set of data entries is configured to cause the computing device to, in accordance with a second security policy, (i) compute an updated confidence score associated with the subset of the set of data entries comprising sensitive data, and (ii) determine whether the updated confidence score exceeds an upper bound of the range of confidence score thresholds; and
based on a determination that the subset of the set of data entries comprise sensitive data as a result of the updated confidence score exceeding the upper bound of the range of confidence score thresholds:
store, via the locally-instantiated application, the subset of the set of data entries in the data storage component as encrypted data entries that satisfy a predefined security requirement.

### Claim 2 (depends on 1)

The client device of claim 1, wherein the processing circuitry is configured to execute the instructions stored in the memory to cause the client device, via instantiation of the application, to process a further subset of the set of data entries in accordance with the first security policy processing operations when the updated confidence score is within the range of confidence score thresholds.

### Claim 3 (depends on 1)

The client device of claim 1, wherein the processing circuitry is configured to execute the instructions stored in the memory to cause the client device to determine the location of the set of data entries stored in the data storage component in response to one or more triggering conditions being met, and
wherein the one or more triggering conditions comprise one or more of a time of day, CPU usage being less than a threshold CPU usage, or memory usage being less than a threshold memory usage.

### Claim 4 (depends on 1)

The client device of claim 1, wherein the processing of the subset of the set of data entries in accordance with the first security policy processing operations comprises executing a machine learning algorithm to determine whether each one of the subset of the set of data entries matches a predefined data format identified with sensitive data.

### Claim 5 (depends on 4)

The client device of claim 4, wherein the processing of the subset of the set of data entries in accordance with the first security policy processing operations comprises using stored metadata as feedback to the machine learning algorithm that indicates which data entries were previously identified as comprising sensitive or non-sensitive data.

### Claim 6 (depends on 1)

The client device of claim 1, wherein the computing the updated confidence score in accordance with the second security policy comprises executing a lightweight machine learning algorithm, which is trained via training data comprising sample data entries that are labeled with an indication of whether each training data sample satisfies a predefined data structure, to determine whether each one of the subset of the set of data entries matches a predefined sensitive data type structure.

### Claim 7 (depends on 1)

The client device of claim 1, wherein the processing circuitry is configured to execute the instructions stored in the memory to cause the client device to process the subset of the set of data entries in accordance with a third security policy that comprises verifying a validity of the subset of the set of data entries based upon an identified data type.

### Claim 8

A computer-implemented method, comprising:
causing local instantiation of an application on a client device that is coupled to a plurality of data storage components that form part of a data network, the application being configured to (i) enable the client device to access data entries from the plurality of data storage components, (ii) obtain a location of a set of data entries stored in a data storage component from among the plurality of data storage components, and (iii) determine whether stored data entries comprises sensitive or non-sensitive data by:
processing, using first security policy processing operations that are computationally offloaded from a computing device in the data network, a subset of the set of data entries to compute a confidence score associated with the subset of the set of data entries comprising sensitive data;
based on a determination that the confidence score is within a range of confidence score thresholds:
transmitting at least a portion of the subset of the set of data entries to the computing device as encrypted and tokenized data,
wherein transmitting the at least the portion of the subset of the set of data entries is configured to cause the computing device to, in accordance with a second security policy, (i) compute an updated confidence score associated with the subset of the set of data entries comprising sensitive data, and (ii) determine whether the updated confidence score exceeds an upper bound of the range of confidence score thresholds; and
based on a determination that the subset of the set of data entries comprise sensitive data as a result of the updated confidence score exceeding the upper bound of the range of confidence score thresholds:
storing the subset of the set of data entries in the data storage component as encrypted data entries that satisfy a predefined security requirement.

### Claim 9 (depends on 8)

The computer-implemented method of claim 8, further comprising:
processing a further subset of the set of data entries in accordance with the first security policy processing operations when the updated confidence score is within the range of confidence score thresholds.

### Claim 10 (depends on 8)

The computer-implemented method of claim 8, wherein the processing of the subset of the set of data entries in accordance with the first security policy processing operations comprises executing a machine learning algorithm to determine whether each one of the subset of the set of data entries matches a predefined data format identified with sensitive data.

### Claim 11 (depends on 10)

The computer-implemented method of claim 10, wherein the processing of the subset of the set of data entries in accordance with the first security policy processing operations comprises using stored metadata as feedback to the machine learning algorithm that indicates which data entries were previously identified as comprising sensitive or non-sensitive data.

### Claim 12 (depends on 8)

The computer-implemented method of claim 8, wherein the computing the updated confidence score in accordance with the second security policy comprises executing a lightweight machine learning algorithm, which is trained via training data comprising sample data entries that are labeled with an indication of whether each training data sample satisfies a predefined data structure, to determine whether each one of the subset of the set of data entries matches a predefined sensitive data type structure.

### Claim 13 (depends on 8)

The computer-implemented method of claim 8, further comprising:
processing the subset of the set of data entries in accordance with a third security policy that comprises verifying a validity of the subset of the set of data entries based upon an identified data type.

### Claim 14

A non-transitory computer-readable medium having instructions stored thereon that, when executed by processing circuitry of a client device coupled to a plurality of data storage components that form part of a data network, cause:
local instantiation of an application that is configured to enable the client device to access data entries from the plurality of data storage components and to determine whether stored data entries comprises sensitive or non-sensitive data by:
obtaining a location of a set of data entries stored in a data storage component from among the plurality of data storage components;
processing, using first security policy processing operations that are computationally offloaded from a computing device in the data network, a subset of the set of data entries to compute a confidence score associated with the subset of the set of data entries comprising sensitive data;
based on a determination that the confidence score is within a range of confidence score thresholds:
transmitting at least a portion of the subset of the set of data entries to the computing device as encrypted and tokenized data,
wherein transmitting the at least the portion of the subset of the set of data entries is configured to cause the computing device to, in accordance with a second security policy, (i) compute an updated confidence score associated with the subset of the set of data entries comprising sensitive data, and (ii) determine whether the updated confidence score exceeds an upper bound of the range of confidence score thresholds; and
based on a determination that the subset of the set of data entries comprise sensitive data as a result of the updated confidence score exceeding the upper bound of the range of confidence score thresholds:
storing the subset of the set of data entries in the data storage component as encrypted data entries that satisfy a predefined security requirement.

### Claim 15 (depends on 14)

The non-transitory computer-readable medium of claim 14, wherein the instructions, when executed by the processing circuitry, further cause the client device to process a further subset of the set of data entries in accordance with the first security policy processing operations when the updated confidence score is within the range of confidence score thresholds.

### Claim 16 (depends on 14)

The non-transitory computer-readable medium of claim 14, wherein the instructions, when executed by the processing circuitry, further cause the client device to determine the location of the set of data entries stored in the data storage component in response to one or more triggering conditions being met,
wherein the one or more triggering conditions comprise one or more of a time of day, CPU usage being less than a threshold CPU usage value, or a memory usage being less than a memory usage threshold.

### Claim 17 (depends on 14)

The non-transitory computer-readable medium of claim 14, wherein the instructions, when executed by the processing circuitry, cause the client device to process the subset of the set of data entries in accordance with the first security policy processing operations by executing a machine learning algorithm to determine whether each one of the subset of the set of data entries matches a predefined data format identified with sensitive data.

### Claim 18 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the instructions, when executed by the processing circuitry, further cause the client device to store metadata that indicates which data entries were previously identified as comprising sensitive or non-sensitive data, and to use the stored metadata as feedback to the machine learning algorithm.

### Claim 19 (depends on 14)

The non-transitory computer-readable medium of claim 14, wherein the instructions, when executed by the processing circuitry, cause the client device to compute the updated confidence score in accordance with the second security policy by executing a lightweight machine learning algorithm, which is trained via training data comprising sample data entries that are labeled with an indication of whether each training data sample satisfies a predefined data structure, to determine whether each one of the subset of the set of data entries matches a predefined sensitive data type structure.

### Claim 20 (depends on 14)

The non-transitory computer-readable medium of claim 14, wherein the instructions, when executed by the processing circuitry, cause the client device to process the subset of the set of data entries in accordance with a third security policy that comprises verifying a validity of the subset of the set of data entries based upon an identified data type.
