US-12619901-B2 - Intelligently aggregating data using a convolutional network quantum processor

US12619901B2US 12619901 B2US12619901 B2US 12619901B2US-12619901-B2

Abstract

Aspects of the disclosure relate to intelligently aggregating data using a convolutional network quantum processor. Using an artificial intelligence model, a computing platform may identify dependencies derived from a consumer request and corresponding values. For each configuration of dependency and corresponding value, the computing platform may determine a likelihood of success. The computing platform may alter the configurations and may determine a likelihood of success for each altered configuration. The computing platform may train a convolutional network to analyze the consumer request using the configurations, the altered configurations, and/or corresponding likelihoods of success. The convolutional network may generate a function that describes the dependencies derived from the consumer request. Using a quantum processing model, the computing platform may analyze different states of the function in parallel and may determine whether to approve or deny the consumer request based on the analysis.

Inventors

Partha Sarathi Dhar
Ravi Kiran Hukmani
Pratikkumar Dharnendrakumar Shah
Kamal Joshi
Venugopal Ramini
Swarn Deep

Assignees

BANK OF AMERICA CORPORATION

Dates

Publication Date: 20260505
Application Date: 20221118

Claims (20)

1 . A method comprising: at a computing device including one or more processors and memory: receiving, from a consumer computing device, a request to initiate a transaction with an enterprise organization service; transmitting, to the consumer computing device and based on receiving the consumer request, a request for consumer data; receiving, from the consumer computing device, the requested consumer data; analyzing the consumer request to identify one or more dependencies that correspond to the consumer request; identifying, based on analyzing the consumer data, values that correspond to the dependencies; determining a likelihood of success of the transaction based on each combination of a dependency and a corresponding value; identifying, by a machine learning engine, patterns in customer request data within a consumer data database; continuously altering, by the machine learning engine, a configuration of the identified one or more dependencies; determining, by the machine learning engine, a likelihood of success of each altered configuration of the identified one or more dependencies; comparing the altered configuration of the one or more dependencies and associated likelihoods of success to the likelihood of success determined based on each combination of the dependency and the corresponding value; identifying, based on the comparing and by the machine learning engine, patterns that influence the likelihood of success; pooling, by the machine learning engine, the identified patterns that influence the likelihood of success; updating, by the machine learning engine and based on the pooled patterns, the combinations and the determined likelihoods of success, a training dataset; generating, by the machine learning engine and using the updated training dataset, a convolutional network; determining, by an artificial intelligence model and using the convolutional network, a function that describes a relationship between the dependencies; analyzing, using a quantum processing model, the function to identify a configuration of dependencies that results in a maximum likelihood of success of the transaction; comparing the maximum likelihood of success to a satisfaction threshold value; based on determining the maximum likelihood of success satisfies a satisfaction threshold, approving the consumer request; and transmitting, to the consumer computing device, a notification indicating approval of the consumer request.
2 . The method of claim 1 , wherein the request for consumer data further comprises a request for at least one of: personal identifiable information; a financial history; or a current financial status.
3 . The method of claim 1 , wherein the analyzing the consumer request to identify one or more dependencies further comprises: parsing the consumer request using at least one natural language processing (NLP) algorithm; and identifying, based on the parsing, at least one of: a feature that increases the likelihood of success of the transaction, or a feature that decreases the likelihood of success of the transaction.
4 . The method of claim 1 , wherein the determining the likelihood of success of the transaction further comprises: parsing historical data, wherein the historical data comprises at least one of: previously analyzed consumer requests, consumer data that corresponds to the previously analyzed consumer requests, a likelihood of success that corresponds to each previously analyzed consumer request, or a determination that corresponds to each previously analyzed consumer request and indicates whether the previously analyzed consumer requests were approved; comparing the consumer request to the historical data; and determining, based on the comparing, whether the consumer request is similar to at least one previously analyzed consumer request.
5 . The method of claim 4 , further comprising, based on determining the consumer request is similar to the at least one previously analyzed consumer request: comparing the consumer data to the historical data; based on determining the consumer data is similar to the historical data, adjusting the likelihood of success that corresponds to the at least one previously analyzed consumer request based on identifying differences between the dependencies derived from the consumer request and dependencies associated with the at least one previously analyzed consumer request; and determining the likelihood of success of the consumer request based on the adjusting.
6 . The method of claim 5 , further comprising, based on determining the consumer data is not similar to the historical data, determining the likelihood of success based on weighted consumer data.
7 . The method of claim 4 , further comprising, based on determining the consumer request is not similar to the at least one previously analyzed consumer request, determining the likelihood of success based on weighted consumer data.
8 . The method of claim 1 , wherein the analyzing the function further comprises determining, using at least one quantum processing gate, a range of output values from the function, wherein each output value corresponds to a different configuration of dependencies, wherein each dependency within the different configurations of dependencies corresponds to different values of the consumer data, wherein each output value indicates a likelihood of success of the transaction based on the different values of the consumer data that corresponds to the different configurations of dependencies, and wherein the range of output values comprises the maximum likelihood of success.
9 . The method of claim 8 , further comprising determining an accuracy of the maximum likelihood of success, wherein the determining the accuracy comprises: comparing the different configurations of dependencies and the corresponding different values of the consumer data to the training dataset; determining whether dependencies within the different configurations of dependencies correspond to the dependencies within the training dataset; based on determining at least one dependency within at least one different configuration of dependencies does not correspond to the dependencies within the training dataset, determining the maximum likelihood of success is inaccurate; and based on determining that the maximum likelihood of success is inaccurate, instructing the artificial intelligence model to analyze the convolutional network and the training dataset.
10 . The method of claim 9 , further comprising, based on determining the dependencies within the different configurations of dependencies correspond to the dependencies within the training dataset, determining the maximum likelihood of success is accurate.
11 . The method of claim 1 , wherein the satisfaction threshold value indicates a minimum likelihood of success that is needed to approve the consumer request.
12 . The method of claim 11 , wherein the determining the maximum likelihood of success satisfies the satisfaction threshold comprises determining the maximum likelihood of success is equal to or greater than the satisfaction threshold value.
13 . The method of claim 11 , further comprising determining the maximum likelihood of success does not satisfy the satisfaction threshold comprises determining the maximum likelihood of success is less than the satisfaction threshold value.
14 . The method of claim 13 , further comprising, based on determining the maximum likelihood of success does not satisfy the satisfaction threshold, denying the consumer request.
15 . The method of claim 14 , further comprising, based on denying the consumer request, transmitting, to the consumer computing device, a notification indicating denial of the consumer request.
16 . A computing platform comprising: at least one processor; a communication interface communicatively coupled to the at least one processor; and memory storing computer-readable instructions that, when executed by the at least one processor, cause the computing platform to: receive, from a consumer computing device, a request to initiate a transaction with an enterprise organization service; transmit, to the consumer computing device and based on receiving the consumer request, a request for consumer data; receive, from the consumer computing device, the requested consumer data; analyze the consumer request to identify one or more dependencies that correspond to the consumer request; identify, based on analyzing the consumer data, values that correspond to the dependencies; determine a likelihood of success of the transaction based on each combination of a dependency and a corresponding value; identify, by a machine learning engine, patterns in customer request data within a consumer data database; continuously alter, by the machine learning engine, a configuration of the identified one or more dependencies; determine, by the machine learning engine, a likelihood of success of each altered configuration of the identified one or more dependencies; compare the altered configuration of the one or more dependencies and associated likelihoods of success to the likelihood of success determined based on each combination of the dependency and the corresponding value; identify, based on the comparing and by the machine learning engine, patterns that influence the likelihood of success; pool, by the machine learning engine, the identified patterns that influence the likelihood of success; update, by the machine learning engine and based on the pooled patterns, the combinations and the determined likelihoods of success, a training dataset; generate, by the machine learning engine and using the updated training dataset, a convolutional network; determine, by an artificial intelligence model and using the convolutional network, a function that describes a relationship between the dependencies; analyze, using a quantum processing model, the function to identify a configuration of dependencies that results in a maximum likelihood of success of the transaction; compare the maximum likelihood of success to a satisfaction threshold value; based on determining the maximum likelihood of success satisfies a satisfaction threshold, approve the consumer request; and transmit, to the consumer computing device, a notification indicating approval of the consumer request.
17 . The computing platform of claim 16 , wherein the analyzing the consumer request to identify one or more dependencies further causes the computing platform to: parse the consumer request using at least one natural language processing (NLP) algorithm; and identifying, based on the parsing, at least one of: a feature that increases the likelihood of success of the transaction, or a feature that decreases the likelihood of success of the transaction.
18 . The computing platform of claim 16 , wherein the determining the likelihood of success of the transaction further causes the computing platform to: parse historical data, wherein the historical data comprises at least one of: previously analyzed consumer requests, consumer data that corresponds to the previously analyzed consumer requests, a likelihood of success that corresponds to each previously analyzed consumer request, or a determination that corresponds to each previously analyzed consumer request and indicates whether the previously analyzed consumer requests were approved; compare the consumer request to the historical data; and determine, based on the comparing, whether the consumer request is similar to at least one previously analyzed consumer request.
19 . One or more non-transitory computer-readable media storing instructions that, when executed by a computing platform comprising at least one processor, memory, and a communication interface, cause the computing platform to: receive, from a consumer computing device, a request to initiate a transaction with an enterprise organization service; transmit, to the consumer computing device and based on receiving the consumer request, a request for consumer data; receive, from the consumer computing device, the requested consumer data; analyze the consumer request to identify one or more dependencies that correspond to the consumer request; identify, based on analyzing the consumer data, values that correspond to the dependencies; determine a likelihood of success of the transaction based on each combination of a dependency and a corresponding value; identify, by a machine learning engine, patterns in customer request data within a consumer data database; continuously alter, by the machine learning engine, a configuration of the identified one or more dependencies; determine, by the machine learning engine, a likelihood of success of each altered configuration of the identified one or more dependencies; compare the altered configuration of the one or more dependencies and associated likelihoods of success to the likelihood of success determined based on each combination of the dependency and the corresponding value; identify, based on the comparing and by the machine learning engine, patterns that influence the likelihood of success; pool, by the machine learning engine, the identified patterns that influence the likelihood of success; update, by the machine learning engine and based on the pooled patterns, the combinations, and the determined likelihoods of success, a training dataset; generate, by the machine learning engine and using the updated training dataset, a convolutional network; determine, by an artificial intelligence model and using the convolutional network, a function that describes a relationship between the dependencies; analyze, using a quantum processing model, the function to identify a configuration of dependencies that results in a maximum likelihood of success of the transaction; compare the maximum likelihood of success to a satisfaction threshold value; based on determining the maximum likelihood of success satisfies a satisfaction threshold, approve the consumer request; and transmit, to the consumer computing device, a notification indicating approval of the consumer request.
20 . The non-transitory computer-readable media of claim 19 , wherein the satisfaction threshold value indicates a minimum likelihood of success that is needed to approve the consumer request.

Description

BACKGROUND Aspects of the disclosure relate to hardware and software for intelligently aggregating data using a convolutional network quantum processor. In particular, one or more aspects of the disclosure may relate to training an artificial intelligence engine to analyze a consumer request and consumer data, determining a function that describes a relationship between dependencies within the consumer request, training a quantum processing model to analyze the function to determine a configuration of dependencies and corresponding consumer data values that may yield a maximum likelihood of success of the consumer request, and determining whether to approve or deny the consumer request based on the analysis by the quantum processing model. Current data aggregation procedures associated with analyzing consumer transaction requests may require an enterprise organization to gather information from a consumer. Consequently, the enterprise organization may generate and transmit, to the consumer, a plurality of data requests (e.g., requests for personal identifiable information that describes the consumer, a financial history that corresponds to the consumer, a current financial status that corresponds to the consumer, or the like) in response to receiving a consumer request to interact with at least one enterprise organization service (e.g., submit a business acquisition application, inquire about consumer credit, or the like). The enterprise organization may sequentially process the consumer data to analyze the consumer request and to determine whether to allow or deny the consumer request. The enterprise organization may analyze each individual piece of consumer data prior to considering the consumer data as a whole. As such, the efficiency of current data aggregation procedures, and the efficiency of determining whether to approve or deny the consumer request, may be compromised due to a volume of consumer data to be processed. Furthermore, under current data aggregation procedures, the enterprise organization might not maintain a consumer data database and/or a training dataset (e.g., the enterprise organization might not process current consumer requests using previously processed consumer requests that may be similar to current consumer requests, using previously analyzed consumer data that may be similar to current consumer data, and/or using determinations that correspond to the previously processed consumer requests). As such, in instances where the consumer transmits a plurality of requests to utilize a plurality of enterprise organization services, the consumer may receive, from the enterprise organization, a plurality of requests for consumer data. Consequently, the consumer may wait an extended period of time to receive a determination from the enterprise organization indicating approval or denial of each consumer request. Therefore, current data aggregation procedures might not offer the enterprise organization a method for streamlining communication with the consumer, intelligently aggregating consumer data, using an artificial intelligence model and parallel processing to analyze the consumer data, and/or using quantum processing logic to analyze a plurality of consumer requests. SUMMARY The following presents a simplified summary in order to provide a basic understanding of some aspects of the disclosure. The summary is not an extensive overview of the disclosure. It is neither intended to identify key or critical elements of the disclosure nor to delineate the scope of the disclosure. The following summary merely presents some concepts of the disclosure in a simplified form as a prelude to the description below. Aspects of the disclosure provide effective, efficient, and convenient technical solutions that address and overcome the technical problems associated with intelligently aggregating, in real-time or near real-time, data using a convolutional network quantum processor. In accordance with one or more embodiments, a method may comprise, at a computing device including one or more processors and memory, receiving, from a consumer computing device, a request to initiate a transaction with an enterprise organization service. The method may comprise transmitting, to the consumer computing device and based on receiving the consumer request, a request for consumer data. The method may comprise receiving, from the consumer computing device, the requested consumer data. The method may comprise analyzing the consumer request to identify one or more dependencies that correspond to the consumer request. The method may comprise identifying, based on analyzing the consumer data, values that correspond to the dependencies. The method may comprise determining a likelihood of success of the transaction based on each combination of a dependency and a corresponding value. The method may comprise updating a training dataset using the combinations and the determined likelihoods of success. The method may comp