Patent Issued for System to facilitate predictive analytic algorithm deployment in an enterprise (USPTO 11461706): Hartford Fire Insurance Company

2022 OCT 26 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- According to news reporting originating from Alexandria, Virginia, by NewsRx journalists, a patent by the inventors Barnes, Shane Eric (Avon, CT, US), Kapushinski, Tyler L. (New Britain, CT, US), Madison, James A. (Windsor, CT, US), Subramaniam, Mohan (Southington, CT, US), Wasik, Ludwig Steven (West Hartford, CT, US), filed on April 8, 2020, was published online on October 4, 2022.

The assignee for this patent, patent number 11461706, is Hartford Fire Insurance Company (Hartford, Connecticut, United States).

Reporters obtained the following quote from the background information supplied by the inventors: “In some cases, an enterprise might want to analyze, model, and/or predict performance values. For example, a business might want to predict a likelihood of a future event occurring based on a number of different factors. Typically, a user associated with the enterprise may manually define rules and/or logic to implement such predictions. These rules and/or logic can then be tested before being deployed for use by the enterprise. Such an approach, however, can be a time consuming and error-prone process-especially when the logic being implemented for an algorithm is complex and/or a substantial number of factors are associated with the prediction.

“It would be desirable to provide systems and methods to accurately and efficiently facilitate predictive analytic algorithm deployment for an enterprise, while allowing for flexibility and effectiveness when creating, reviewing, and/or monitoring algorithms as appropriate.”

In addition to obtaining background information on this patent, NewsRx editors also obtained the inventors’ summary information for this patent: “According to some embodiments, systems, methods, apparatus, computer program code and means are provided to facilitate predictive analytic algorithm deployment for an enterprise. In some embodiments, an analytics computing environment data store may contain a set of electronic data records, each electronic data record being associated with a predictive analytic algorithm and including an algorithm identifier and a set of algorithm characteristic values. An analytics environment computer may receive an adjustment from a user associated with an enterprise, the adjustment changing at least one of the set of algorithm characteristic values for a predictive analytic algorithm. Deployment of the predictive analytic algorithm may then be initiated in an enterprise operations workflow and at least one result may be generated. The deployed predictive analytic algorithm may then monitor the result and generate an alert signal when the result exceeds a boundary condition.

“Some embodiments comprise: means for receiving, by an analytics environment computer, data from an analytics computing environment data store containing a set of electronic data records, each electronic data record being associated with a predictive analytic algorithm and including an algorithm identifier and a set of algorithm characteristic values; means for receiving, by the analytics environment computer, an adjustment to at least one of the set of algorithm characteristic values from a user associated with the enterprise; means for initiating deployment of the predictive analytic algorithm in an enterprise operations workflow; and means for executing, by an enterprise operations workflow computer, an operations workflow in association with the deployed predictive analytic algorithm to generate at least one result, wherein the deployed predictive analytic algorithm is to monitor the result and generate an alert signal when the result exceeds a boundary condition.

“In some embodiments, a communication device associated with a back-end application computer server exchanges information with remote devices. The information may be exchanged, for example, via public and/or proprietary communication networks.

“Some technical effects of some embodiments of the invention are improved and computerized ways to accurately and efficiently facilitate predictive algorithm deployment for an enterprise. With these and other advantages and features that will become hereinafter apparent, a more complete understanding of the nature of the invention can be obtained by referring to the following detailed description and to the drawings appended hereto.”

The claims supplied by the inventors are:

“1. A system to facilitate predictive analytics for an enterprise, comprising: (a) an analytics computing environment data store containing a set of electronic data records, each electronic data record being associated with a predictive analytic algorithm and including an algorithm identifier and a set of algorithm characteristic values, wherein: (i) an experimental predictive analytic algorithm may be promoted to an underwriting production predictive analytic algorithm, and (ii) an underwriting production predictive analytic algorithm may be demoted to an experimental predictive analytic algorithm; (b) an analytics environment computer, coupled to the coupled to the analytics computing environment data store, programmed to: (i) receive an adjustment associated with the enterprise, the adjustment changing at least one of the set of algorithm characteristic values for a predictive analytic algorithm, and (ii) initiate deployment of the predictive analytic algorithm in an enterprise operations workflow; and © an enterprise operations workflow computer, coupled to the analytics environment computer, programmed to: (i) receive the deployment initiation from the analytics environment computer, and (ii) execute an operations workflow in association with the deployed predictive analytic algorithm to generate at least one result, wherein the predictive analytic algorithm stores its own output as an input, allowing for self-tuning while deployed in the enterprise operations workflow computer, wherein the deployed predictive analytic algorithm is to monitor the result and generate an alert signal when the result exceeds a boundary condition, and the alert is further to automatically promote or demote the predictive analytic algorithm as appropriate.

“2. The system of claim 1, wherein the predictive analytic algorithm is associated with a data science algorithm and at least one of the analytics environment computer and the enterprise operations workflow computer are designed to let the data science algorithm to be moved directly to the operations workflow and operate with a level of rigor and reliability associated with the operations workflow.

“3. The system of claim 2, wherein movement of the data science algorithm from the analytics computing environment data store to the operations workflow is performed automatically via system components created for data scientist users.

“4. The system of claim 2, wherein use of the data science algorithm is embedded directly in the operations workflow while simultaneously keeping control of the data science algorithm in a platform associated with an analytics environment computer.

“5. The system of claim 2, wherein components of the system allow for both current underwriting production data science algorithms and current experimental data science algorithms to function in parallel without conflict.

“6. The system of claim 1, wherein system components are constructed with environmental uniformity, reducing overhead associated with movement of system components between the analytics environment computer and the enterprise operations workflow computer.

“7. The system of claim 1, wherein users of the system are configured with environmental uniformity, improving user productivity.

“8. The system of claim 1, wherein user actions are monitored, captured, and reported to achieve a level of legal, privacy, and policy compliance commensurate with a sensitivity of data being processed.

“9. The system of claim 1, wherein the predictive analytic algorithm monitors and checks its own behavior and provides indicators of its own behavior capable of being transmitted to other systems in the enterprise.

“10. The system of claim 1, wherein the predictive analytic algorithm has boundary and error conditions, such that the algorithm will connect to an enterprise logging infrastructure to notify appropriate parties when the algorithm moves outside the boundary condition.

“11. The system of claim 1, wherein a library of reusable components is maintained consistent across data science, operations, current production release, and experimental levels to facilitate movement of data science functionality among the various regions of the enterprise.

“12. The system of claim 1, wherein a model can execute against a regression test base and report a health value as compared to prior baselines.

“13. The system of claim 1, wherein deployment of the predictive analytic algorithm is a self-service process performed by a data scientist in substantially real-time.

“14. A computerized method to automate algorithm deployment for an enterprise, comprising: receiving, by an analytics environment computer, data from an analytics computing environment data store containing a set of electronic data records, each electronic data record being associated with a predictive analytic algorithm and including an algorithm identifier and a set of algorithm characteristic values, wherein: (i) an experimental predictive analytic algorithm may be promoted to an underwriting production predictive analytic algorithm, and (ii) an underwriting production predictive analytic algorithm may be demoted to an experimental predictive analytic algorithm; receiving an adjustment associated with the enterprise, the adjustment changing at least one of the set of algorithm characteristic values for a predictive analytic algorithm; initiating deployment of the predictive analytic algorithm in an enterprise operations workflow; and executing, by an enterprise operations workflow computer, an operations workflow in association with the deployed predictive analytic algorithm to generate at least one result, wherein the predictive analytic algorithm stores its own output as an input, allowing for self-tuning while deployed in the enterprise operations workflow computer, wherein the deployed predictive analytic algorithm is to monitor the result and generate an alert signal when the result exceeds a boundary condition, and the alert is further to automatically promote or demote the predictive analytic algorithm as appropriate.

“15. The method of claim 14, wherein the predictive analytic algorithm is associated with a data science algorithm and at least one of the analytics environment computer and the enterprise operations workflow computer are designed to let the data science algorithm to be moved directly to the operations workflow and operate with a level of rigor and reliability associated with the operations workflow.

“16. The method of claim 15, wherein movement of the data science algorithm from the analytics computing environment data store to the operations workflow is performed automatically via system components created for data scientist users.

“17. The method of claim 15, wherein components of the system allow for both current underwriting production data science algorithms and current experimental data science algorithms to function in parallel without conflict.

“18. The method of claim 14, wherein system components are constructed with environmental uniformity, reducing overhead associated with movement of system components between the analytics environment computer and the enterprise operations workflow computer.

“19. The method of claim 14, wherein users of the system are configured with environmental uniformity, improving user productivity.

“20. The method of claim 14, wherein user actions are monitored, captured, and reported to achieve a level of legal, privacy, and policy compliance commensurate with a sensitivity of data being processed.”

For more information, see this patent: Barnes, Shane Eric. System to facilitate predictive analytic algorithm deployment in an enterprise. U.S. Patent Number 11461706, filed April 8, 2020, and published online on October 4, 2022. Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=11461706.PN.&OS=PN/11461706RS=PN/11461706

(Our reports deliver fact-based news of research and discoveries from around the world.)