Intelligent preprocessing routing to decisioning services

Patent metadata

Field	Value
Publication number	US-2020372575-A1
Application number	US-202016882112-A
Country	US
Kind code	A1
Filing date	May 22, 2020
Priority date	May 23, 2019
Publication date	Nov 26, 2020
Grant date	—

Abstract

The system and methods described herein allow users to give their applicant information when seeking to purchase a good with financing from multiple potential lenders, and may be pre-screened by one or more rule sets implemented by a marketplace client in an eligibility analysis to ultimately submit applicant information to a subset of the multiple potential lenders which are found to be suitable for lending to an applicant based on the applicant information, wherein lender microservices are then run in a jailed, firewalled, and self-contained, autonomous environment, and the results of said lender microservices are reported to the user and may be used to change the one or more rule sets implemented by the marketplace client for future pre-screening of applicants.

First claim

What is claimed is:

1 . A method comprising:
receiving, at a server, through an application user interface, applicant information that is inputted by an applicant and encrypted upon receipt or at a predetermined later time; receiving through another application user interface, one or more rule sets for prequalifying an applicant for a plurality of lenders, wherein the one or more rule sets are stored encrypted in an applicant repository; applying a lender pre-screening rule set to the applicant information to determine a subset of lenders for which the applicant is suitable for borrowing from; receiving, at the server, for the subset of lenders which the applicant is deemed to be suitable for borrowing from, the applicant information and the rule sets for prequalifying the applicant, wherein rule sets for each respective lender out of the subset of lenders and applicant information are decrypted inside of a jailed and non-accessible environment to any user of the server; using an application protocol interface to access the one or more rule sets stored at the server, and applying the rule sets to the applicant information; returning a status value for prequalifying the applicant from the jailed and non-accessible environment; and updating the lender pre-screening rule set based on the status value;

2 . The method of claim 1 , wherein the applicant information includes an external record of a market source which the applicant traversed before reaching the server, and wherein applying of the lender pre-screening rule set further comprises
restricting a plurality of lenders to a further subset of the subset of lenders the applicant is suitable for borrowing from, based on the market source.

3 . The method of claim 2 , further comprising:
applying applicant preferences to the restricted plurality of lenders of the further subset, to still further restrict the lenders based on filters applied by the applicant; displaying restricted lenders; and displaying non-restricted lenders, such that there is a visual distinction between restricted and non-restricted lenders, on a GUI display to the user.

4 . The method of claim 1 , the applying of the lender pre-screening rule set further comprising:
restricting a plurality of lenders to a subset of lenders based on Boolean logic or machine-learning logic applied to the information inputted by the applicant, wherein the updating the lender pre-screening rule set comprises updating Boolean logic operators, applicant information thresholds, or machine learning logic, based on the status value.

5 . The method of claim 1 , wherein applying of the lender pre-screening rule set to determine a subset of lenders for which the applicant is suitable for borrowing from further comprises:
applying applicant information to a lender-specific neural network machine-learning classifier with backpropagation and at least one hidden layer, wherein the user inputted information forms an input layer of the neural network classifier, and the output layer of the classifier contains two nodes, wherein the applicant is deemed to be suitable for borrowing from a particular lender based on the values of the two nodes.

6 . The method of claim 5 , further comprising:
updating the lender pre-screening rule set by adjusting weights between the nodes of the neural network classifier based on the status value for prequalifying the applicant returned from the jailed and non-accessible environment.

7 . The method of claim 1 , further comprising:
receiving, at the server, applicant information for a plurality of marketplace clients, where rules for applying a lender pre-screening rule set for every separate marketplace client are kept in a jailed environment and are encrypted with a marketplace client specific key.

8 . A system comprising:
a memory; a server; a processor coupled to said memory, the processor configured to: receive, at the server, through an application user interface, applicant information that is inputted by an applicant, which is encrypted upon receipt, or encrypted at a predetermined later time, and includes an external record of a market source which the applicant traversed before reaching the server; receive, at the server, through another application user interface, one or more rule sets for prequalifying an applicant for a plurality of lenders, wherein the one or more rule sets are encrypted upon receipt, and stored in an applicant repository; apply a lender pre-screening rule set to the applicant information to determine a subset of lenders for which the applicant is suitable for borrowing from; receive, at the server, for those subset of lenders which the applicant is deemed to be suitable for borrowing from, the applicant information and the rule sets for prequalifying the applicant; decrypt rule sets for each respective lender out of the subset of lenders and applicant information inside of a jailed and non-accessible environment on said server, which is non-accessible to any user of the server, the rule sets and applicant information; use an application protocol interface to access the one or more rule sets stored at the server, and apply the rule sets to the applicant information; and return, to the application user interface, a status value for prequalifying the applicant from the jailed and non-accessible environment;

9 . The system of claim 8 , wherein in applying the lender pre-screening rule set, the processor is further configured to:
restrict a plurality of lenders to a further subset of the subset of lenders the applicant is suitable for borrowing from, based on the market source.

10 . The system of claim 9 , wherein the processor is further configured to:
apply applicant preferences to the restricted plurality of lenders, to still further restrict the lenders based on filters applied by the applicant; display restricted lenders; and display non-restricted lenders, such that there is a visual distinction between restricted and non-restricted lenders, on a GUI display to the user.

11 . The system of claim 8 , wherein the processor is further configured to, in applying the lender pre-screening rule set:
restrict a plurality of lenders to a subset of lenders based on Boolean logic or machine-learning logic applied to the information inputted by the applicant.

12 . The system of claim 8 , wherein the processor, in applying the lender pre-screening rule set to determine a subset of lenders for which the applicant is suitable for borrowing from, is further configured to:
apply applicant information to a lender-specific neural network machine-learning classifier with backpropagation and at least one hidden layer, wherein the user inputted information forms an input layer of the neural network classifier, and the output layer of the classifier contains two nodes, wherein the applicant is deemed to be suitable for borrowing from a particular lender based on the values of the two nodes.

13 . The system of claim 12 , wherein the processor is further configured to:
update the lender pre-screening rule set by adjusting weights between the nodes of the neural network classifier based on the status value for prequalifying the applicant returned from the jailed and non-accessible environment.

14 . The system of claim 8 , wherein the processor is further configured to:
receive, at the server, applicant information for a plurality of marketplace clients, where rules for applying a lender pre-screening rule set for every separate marketplace client are kept in a jailed environment and are encrypted with a marketplace client specific key.

Assignees

• Capital One Services Llc

Inventors

• Kesiboyana Satish
• Vijayaraghavan Rajaboopathy
• Subramanian Nithya
• Sickert Thomas

Classifications

• G06Q40/0305 Primary

  Platforms for credit or lending product research, comparison or matching · CPC title

• G06N3/048

  Activation functions · CPC title

• G06Q40/03

  Credit; Loans; Processing thereof · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06N20/10

  using kernel methods, e.g. support vector machines [SVM] · CPC title