Automated regulatory compliance for insurance

Patent metadata

Field	Value
Publication number	US-11620713-B2
Application number	US-201715725540-A
Country	US
Kind code	B2
Filing date	Oct 5, 2017
Priority date	Aug 22, 2017
Publication date	Apr 4, 2023
Grant date	Apr 4, 2023

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, are described that enable identifying an insurance information associated with a property, retrieving one or more compliance rules related to the insurance information, automatically determining, based on the one or more compliance rules, whether the insurance information has a compliant status or a non-compliant status, in response to determining that the insurance information has the non-compliant status, processing the insurance information, based on one or more exception rules, to generate an exception waiver including a plurality of waiver parameters, applying the plurality of waiver parameters to a waiver model to determine a waiver confidence score, and based on the waiver confidence score selectively generate an insurance certificate for the property.

First claim

What is claimed is:

1. A computer-implemented method executed by one or more processors, the method comprising:
receiving, by one or more servers that are associated with a mortgage-issuing institution that imposes a requirement on borrowers to possess flood protection certificates and that provide real-time status tracking of insurance certificates for mortgaged properties for different kinds of mortgage insurance and from a particular borrower among multiple borrowers that are associated with the mortgage-issuing institution, and through a feed that is associated with the particular borrower, insurance information associated with a property, in a received format as sent by the particular borrower;
standardizing the insurance information associated with the property from the received format into a textual format according to one or more naming conventions;
upon expiration of a scheduled timeline that is associated with the particular borrower, retrieving, by the one or more servers that are associated with the mortgage-issuing institution and from one or more servers that are associated with a government agency, one or more compliance rules related to the standardized insurance information;
automatically determining, by the one or more servers that are associated with the mortgage-issuing institution, that the standardized insurance information in the textual format indicates a non-compliant status for at least one of the one or more compliance rules;
in response to determining that the insurance information indicates the non-compliant status for at least one of the one or more compliance rules, processing, by the one or more servers that are associated with the mortgage-issuing institution, the standardized insurance information using one or more exception rule machine learning models that apply one or more exception rules that define an acceptable discrepancy between two or more terms of the standardized insurance information, to output an exception waiver comprising a plurality of waiver parameters that includes a parameter that identifies the at least one of the one or more compliance rules for which the non-compliant status was automatically determined;
generating, by the one or more servers that are associated with the mortgage-issuing institution and using at least some of the standardized insurance information and the plurality of waiver parameters, a feature matrix that includes data for the standardized insurance information and the plurality of waiver parameters;
providing, by the one or more servers that are associated with the mortgage-issuing institution and to a first machine learning-trained waiver model, the feature matrix that includes data for the standardized insurance information and the plurality of waiver parameters to cause the first machine learning-trained, waiver model to determine a first waiver confidence score for the standardized insurance information that is based on the one or more exception rules and that is associated with the property;
receiving, by the one or more servers that are associated with the mortgage-issuing institution and from the first machine learning-trained waiver model, the first waiver confidence score for the standardized insurance information that is based on the one or more exception rules and that is associated with the property;
determining that the first waiver confidence score fails to satisfy a first predetermined threshold;
only after determining that the first waiver confidence score fails to satisfy the first predetermined threshold, providing, by the one or more servers that are associated with the mortgage-issuing institution and to a different, second machine learning-trained waiver model, the feature matrix that includes data for the standardized insurance information and the plurality of waiver parameters to cause the second machine learning-trained waiver model to determine a second waiver confidence score for the standardized insurance information that is based on the one or more exception rules and that is associated with the property;
receiving, by the one or more servers that are associated with the mortgage-issuing institution and from the second machine-learned waiver model, the second waiver confidence score for the standardized insurance information that is based on the one or more exception rules and that is associated with the property;
based on whether the second waiver confidence score satisfies a threshold confidence score, selectively force-placing, by the one or more servers that are associated with the mortgage-issuing institution, an insurance certificate for the property, and automatically transmitting, by an integrated email module of the one or more servers that are associated with the mortgage-issuing institution and in real time to the expiration of the scheduled timeline, a message indicating that the mortgage institution has force-placed the insurance certificate for the property using a pre-defined letter template;
receiving data identifying a human intervention taken in connection with the output of the one or more exception rule machine learning models;
generating, as training data for retraining the one or more exception rule machine learning models, (i) the data identifying the human intervention taken in connection with the output of the one or more exception rule machine learning models, (ii) the standardized insurance information, and (iii) the plurality of waiver parameters;
retraining the one or more exception rule machine learning models using the training data to apply one or more updated exception rules, to thereby decrease human intervention on exceptions being identified and flagged by the one or more servers that are associated with the mortgage-issuing institution over time; and
using the retrained one or more exception rule machine learning models to process subsequently received, standardized insurance information by applying the one or more updated exception rules.

2. The method of claim 1 , wherein the one or more exception rules comprise at least one of an exception coverage percentage, an exception type, an exception level, an exception term, and an exception priority.

3. The method of claim 1 , wherein force-placing the insurance certificate comprises generating an insurance certificate for the property.

4. The method of claim 1 , wherein the insurance information comprises at least one of a property location, an insurance policy period, an insurance limit, an insurance deductible, and a lapse in insurance coverage.

5. The method of claim 1 , wherein the one or more compliance rules comprise a government law or a government regulation related to the insurance information.

6. The method of claim 5 , wherein determining, based on the one or more compliance rules, whether the insurance information has the compliant status or the non-compliant status comprises:
determining, based on the government law or the government regulation, hazard data that indicates one or more risk zones;
comparing a geographic location of the property with the one or more risk zones;
determining a designated insurance premium or a designated insurance rate for the property based on comparing a geographic location of the property with the risk zones; and
determining whether the insurance information complies with the designated insurance premium or the designated insurance rate for the property.

7. The method of claim 1 , wherein determining that the insurance information has the non-compliant status comprises:
determining that the property does not comply with a requirement to purchase insurance as a result of receiving risk-related financial assistance from a federal government.

8. The method of claim 1 , wherein determining that the insurance information

Assignees

• Accenture Global Solutions Ltd

Inventors

• Ashly Peter
• Rajagopalan Srinivasan Krishnan
• Kamalakannan Venkatesan
• Prakash Anand Yesuraj
• Chandran Manikandan
• Paramasivan Thangaraj
• Jalajam Sivasubramanian Arumugam
• Muthu Loganathan
• Varghese Vinu
• Viswanathan Kumar

Classifications

• G06Q10/10

  Office automation; Time management · CPC title

• G06Q40/08 Primary

  Insurance · CPC title