Granted Patent B1

US 12,602,636 · App. 17/929,469 · Granted Apr 14, 2026

Inventors: Sanjay Kumar Rout (Edison, NJ); Zixuan Yang (New York, NY); Raj Kumar (North Brunswick, NJ)

Assignee: Morgan Stanley Services Group Inc.

G06Q10/06315G06F3/0482

Abstract

Claims (55)

1 . A method for automatically and dynamically configuring and onboarding clients and accounts of a firm, the method comprising:

receiving, by a client and account assembly computer system, from a computer-implemented product loader, a request for the account creation for a of the client, wherein the product loader is configured to receive account-creation requests from multiple channels of the firm for different product types, account types, and channel types, and wherein the request includes information for the account of the client, wherein the information comprises an account type for the account, and wherein the information comprises metadata about the account, wherein the metadata comprise an entity type for the client, a business area of the firm requesting the account, and a channel for the account of the client;

based on the request, generating, by the client and account assembly computer system, an account structure for the account of the client, wherein the account structure comprises applicable account-assembly properties, behaviors and rules based on the metadata and based on a pre-defined master list of properties, behaviors, and rules, wherein generating the account structure comprises:

determining the applicable account-assembly properties, behaviors and rules for the account structure by applying a framework, the framework comprising:

(i) a pre-classification component that dynamically determines account type, eligibility, and rule categories based on the metadata;

(ii) a main processing component that aggregates client profile and demographic data; and

(iii) a post-processing component that validates the profile, performs account grouping, and applies suitability rules using dynamically configurable business rules; and

processing account structure data for the account with a first set of dynamically triggered serial software-implemented filters in a first data pipeline using a pipe-and-filter pattern, wherein:

the data pipeline comprises a plurality of serial software-implemented filters, each filter embodied as an object-oriented class stored in memory of the client and account assembly computer system, the class including instructions that, when executed, update the account structure data of an account object resident in the memory;

each software-implemented filter is mapped to a distinct property defined in a pre-specified master list of properties, the mapping causing the filter to add property-specific configuration data and associated rule data to the account object during execution;

the software-implemented filters are dynamically triggered and reconfigured based on metadata in the client account request, thereby modifying an execution order of the pipeline without redeploying the system; and

a resulting account structure data comprises a mutated account object containing property-specific executable instructions and rule data corresponding to the applicable rules for the account of the client; and

after generating the account structure for the account of the client, generating by the client and account assembly computer system, a final account, wherein generating the final account comprises processing the account structure with a second set of dynamically triggered serial software-implemented filters in a second data pipeline using a pipe-and-filter pattern,

wherein processing the account structure comprises:

validating, by a machine-learned anomaly detection model stored in the memory and trained on historical onboarding outcomes, the account object as it exits the second pipeline, the validating comprising:

monitoring data and stat of the account object as updated by the second set of filters;

reviewing the data and state of the account object for consistency with learned account structure patterns and applicable rules and requirements, such that upon detection of a deviation or orchestration fault, the account object is correctable prior to finalization; and

updating the account structure, upon successful validation, by executing the second set of dynamically triggered serial software-implemented filters arranged in a pipe-and-filter architecture, wherein each dynamically triggered serial software-implemented filter in the second set is implemented as an object-oriented class that, when invoked, adds executable onboarding instructions into the account object, updates onboarding-related data fields of the account object, and passes an updated object sequentially through a pipe to a next filter,

wherein the applicable onboarding processes and services are selected by a rule engine from a pre-specified master list of onboarding processes and services, selection being based on metadata in the client account request specifying the account type, business area, and channel; and

wherein execution of the second set of filters produces a final account object stored in the memory, the final account object comprising a complete, validated data structure containing executable code, service bindings, and configuration data required for immediate use of the account in its specified business area and channel.

2 . The method of claim 1 , wherein the client and account assembly computer system comprises one software-implemented filter in the first set of serial software-implemented filters for each applicable property for the account of the client, such that the one software-implemented filter for each applicable property processes the account structure data to update the account structure data to implement at least one applicable rule for the applicable property for the account of the client.

3 . The method of claim 2 , wherein the client and account assembly computer system comprises one software-implemented filter in the second set of serial software-implemented filters for each applicable onboarding process for the account of the client, such that the one software-implemented filter for each applicable onboarding process processes the account structure to update the account structure with data from the onboarding process for the account of the client.

4 . The method of claim 1 , wherein pipes for the first and second pipe-and-filter data pipelines comprise an objected oriented class to trigger a current software-implemented filter in the data pipeline, pass an output of the current software-implemented filter as input to a next software-implemented filter in the data pipeline, and trigger the next software-implemented filter in the data pipeline.

5 . The method of claim 1 , further comprising:

prior to generating the final account object, training, by the client and account assembly computer system, through machine learning, multiple machine learning systems to validate the final account object; and

after training the multiple machine learning systems, validating, by the multiple machine learning systems, the final account object.

6 . The method of claim 1 , wherein the account type is an investment account and the firm is a wealth management firm.

7 . A computer system for automatically and dynamically configuring and onboarding clients and accounts of a firm, the computer system comprising:

one or more processors; and

a memory in communication with the one or more processors, wherein the memory stores instructions that, when executed by the one or more processors, cause the one or more processors to:

receive, from a computer-implemented product loader, a request for account creation for a client, wherein the request includes information for the account of the client, wherein the product loader is configured to receive account-creation requests from multiple channels of the firm for different product types, account types, and channel types, and wherein the information comprises an account type for the account, and wherein the information comprises metadata about the account, wherein the metadata comprise an entity type for the client, a business area of the firm requesting the account, and a channel for the account;

based on the request, generate an account structure for the account of the client, wherein the account structure comprises applicable account-assembly properties, behaviors and rules based on the metadata and based on a pre-defined master list of properties, behaviors and rules, wherein the one or more processors are configured to generate the account structure by:

determining the applicable account-assembly properties, behaviors and rules for the account structure by applying a framework, the framework comprising:

(i) a pre-classification component that dynamically determines account type, eligibility, and rule categories based on the metadata;

(ii) a main processing component that aggregates client profile and demographic data; and

(iii) a post-processing component that validates the profile, performs account grouping, and applies suitability rules using dynamically configurable business rules; and

processing account structure data for the account with a first set of serial software-implemented filter that are dynamically triggered in a first data pipeline a using pipe-and-filter pattern, wherein;

the data pipeline comprises a plurality of serial software-implemented filters, each filter embodied as an object-oriented class stored in memory of the client and account assembly computer system, the class including instructions that, when executed, update the account structure data of an account object resident in the memory:

each software-implemented filter is mapped to a distinct property defined in a pre-specified master list of properties, the mapping causing the filter to add property-specific configuration data and associated rule data to the account object during execution;

the software-implemented filters are dynamically triggered and reconfigured based on metadata in the client account request, thereby modifying an execution order of the pipeline without redeploying the system; and

a resulting account structure data comprises a mutated account object containing property-specific executable instructions and rule data corresponding to the applicable rules for the account of the client; and

after generating the account structure for the account, generate a final account, wherein generating the final account comprises processing the account structure with a second set of serial software-implemented filters that are dynamically triggered in a second data pipeline using a pipe-and-filter pattern, wherein processing the account structure comprises:

validating, by a machine-learned anomaly detection model stored in the memory and trained on historical onboarding outcomes, the account object as it exits the second pipeline, the validating comprising:

monitoring data and stat of the account object as updated by the second set of filters;

reviewing the data and state of the account object for consistency with learned account structure patterns and applicable rules and requirements, such that upon detection of a deviation or orchestration fault, the account object is correctable prior to finalization; and

updating the account structure, upon successful validation, by executing the second set of dynamically triggered serial software-implemented filters arranged in a pipe-and-filter architecture, wherein each dynamically triggered serial software-implemented filter in the second set is implemented as an object-oriented class that, when invoked, adds executable onboarding instructions into the account object, updates onboarding-related data fields of the account object, and passes an updated object sequentially through a pipe to a next filter,

wherein the applicable onboarding processes and services are selected by a rule engine from a pre-specified master list of onboarding processes and services, selection being based on metadata in the client account request specifying the account type, business area, and channel; and

wherein execution of the second set of filters produces a final account object stored in the memory, the final account object comprising a complete, validated data structure containing executable code, service bindings, and configuration data required for immediate use of the account in its specified business area and channel.

8 . The computer system of claim 7 , wherein the computer system comprises one software-implemented filter in the first set of serial software-implemented filters for each applicable property for the account of the client, such that the one software-implemented filter for each applicable property processes the account structure data to update the account structure data to implement at least one applicable rule for the applicable property for the account of the client.

9 . The computer system of claim 8 , wherein the computer comprises system one software-implemented filter in the second set of serial software-implemented filters for each applicable onboarding process for the account of the client, such that the one software-implemented filter for each applicable onboarding process processes the account structure to update the account with data from the onboarding process for the account of the client.

10 . The computer system of claim 7 , wherein pipes for the first and second pipe-and-filter data pipelines comprise an objected oriented class to trigger a current software-implemented filter in the data pipeline, pass an output of the current software-implemented filter as input to a next software-implemented filter in the data pipeline, and trigger the next filter in the data pipeline.

11 . The computer system of claim 7 , wherein the memory stores instructions that when executed by the one or more processors cause the one or more processors to:

prior to generating the final account object, train, through machine learning, multiple machine learning systems to validate the final account object; and

after training the multiple machine learning systems, validate the final account object with the multiple machine learning systems.

12 . The computer system of claim 7 , wherein the account type is an investment account and the firm is a wealth management firm.

Assignments (1)

References Cited (33)