Computer-based systems and methods configured to utilize automating deployment of predictive models for machine learning tasks

Patent metadata

Field	Value
Publication number	US-12461728-B2
Application number	US-202218059091-A
Country	US
Kind code	B2
Filing date	Nov 28, 2022
Priority date	Jul 12, 2019
Publication date	Nov 4, 2025
Grant date	Nov 4, 2025

Abstract

A method includes obtaining feature generation code from, which is configured to determine features relating to input data. The method further includes obtaining data grouping code, which is configured to generate training data by determining a plurality of data groupings for the features relating to the input data. The method further includes obtaining modeling code, which is derived at least in part by applying one or more machine learning algorithms to the training data. The method further includes applying a model wrapper code to the feature generation code, the data grouping code, and the modeling code to generate a model wrapper and deploying the model wrapper such that the model wrapper may receive a first application programming interface (API) call including an input data value, determine a score relating to the input data value, and send a second API call including the score in response to the first API call.

First claim

What is claimed is:

1 . A system comprising:
a memory; at least one processor coupled to the memory; and non-transitory, computer readable instructions stored on the memory that, upon execution, cause the system to:
apply a model wrapper code to a plurality of components to generate a model wrapper, wherein at least one of the plurality of components comprises a data grouping code comprising first instructions for categorizing data analyzed by the model wrapper into one or more of a first plurality of data groups;
deploy the model wrapper to categorize the data into the one or more of the first plurality of data groups;
receive or obtain, after the model wrapper code was applied to generate the model wrapper and after the model wrapper was deployed, at least one updated component for the model wrapper, wherein the at least one updated component comprises an updated data grouping code for the model wrapper, wherein the updated data grouping code comprises second instructions for categorizing further data analyzed by the model wrapper into one or more of a second plurality of data groups that is different from the first plurality of data groups;
re-apply the model wrapper code to generate an updated model wrapper, wherein the model wrapper is still in use while the model wrapper code is being re-applied;
apply an application programming interface (API) configurator to the updated model wrapper, wherein the API configurator is configured to, upon application of the API configurator to the updated model wrapper, cause the updated model wrapper to receive and respond to predetermined API calls after the updated model wrapper is deployed, wherein the predetermined API calls are defined by the API configurator; and
deploy the updated model wrapper to incorporate the at least one updated component into the updated model wrapper.

2 . The system of claim 1 , wherein the at least one updated component for the model wrapper comprises updated code for the model wrapper, and wherein the updated code comprises an updated feature generation code for the model wrapper, wherein the updated feature generation code is configured to determine features relating to input data.

3 . The system of claim 1 , wherein the updated data grouping code is further configured to generate training data by determining a plurality of data groupings for features relating to input data.

4 . The system of claim 1 , wherein the at least one updated component for the model wrapper comprises updated code for the model wrapper, and wherein the updated code comprises an updated modeling code for the model wrapper, wherein the updated modeling code is derived at least in part by applying one or more machine learning algorithms to training data.

5 . The system of claim 1 , wherein the at least one updated component for the model wrapper does not affect or change the model wrapper until after the updated model wrapper is generated using the at least one updated component.

6 . The system of claim 1 , wherein the at least one updated component for the model wrapper is stored as an artifact in a repository.

7 . The system of claim 6 , wherein an update to the artifact in the repository does not affect or change the model wrapper until after the updated model wrapper is generated using the artifact in the repository.

8 . The system of claim 1 , wherein the model wrapper comprises feature generation code, the data grouping code, and modeling code.

9 . The system of claim 8 , wherein the computer readable instructions further causes the system to form the model wrapper based on an application of the model wrapper code to the feature generation code, the data grouping code, and the modeling code.

10 . The system of claim 9 , wherein the computer readable instructions further causes the system to deploy the model wrapper, wherein deployment of the model wrapper configures the model wrapper to receive an application programming interface (API) call comprising an input data value.

11 . The system of claim 10 , wherein the API call is a first API call, and wherein the deployment of the model wrapper further configures the model wrapper to determine a score relating to the input data value and send a second API call comprising the score in response to the first API call.

12 . The system of claim 11 , wherein the computer readable instructions further causes the system to deploy the updated model wrapper.

13 . The system of claim 12 , wherein the input data value is a first input data value and the score is a first score, and wherein the deployment of the updated model wrapper configures the updated model wrapper to:
receive a third API call comprising a second input data value; determine a second score relating to the second input data value; and send a fourth API call comprising the second score in response to the third API call.

14 . A method comprising:
determining, by one or more processors of a computing device, a model wrapper for re-deployment, wherein:
the model wrapper has been previously deployed for use, and
the model wrapper comprises a plurality of components, wherein at least one of the plurality of components comprises a data grouping code comprising first instructions for categorizing data analyzed by the model wrapper into one or more of a first plurality of data groups;
receiving or obtaining, by the one or more processors, at least one updated component for the model wrapper, wherein the at least one updated component comprises an updated data grouping code for the model wrapper, wherein the updated data grouping code comprises second instructions for categorizing the data analyzed by the model wrapper into one or more of a second plurality of data groups that is different from the first plurality of data groups; generating, by the one or more processors, an updated model wrapper using the model wrapper and the at least one updated component, wherein the generating of the updated model wrapper occurs while the model wrapper is being currently deployed and in use; and applying an application programming interface (API) configurator to the updated model wrapper, wherein the API configurator is configured to, upon application of the API configurator to the updated model wrapper, cause the updated model wrapper to receive and respond to predetermined API calls after the updated model wrapper is deployed, wherein the predetermined API calls are defined by the API configurator.

15 . The method of claim 14 , wherein the determining of the model wrapper for re-deployment is based on determining that the at least one updated component for the model wrapper is available.

16 . The method of claim 15 , wherein the at least one updated component for the model wrapper is available in a code repository related to the model wrapper, and further wherein the at least one updated component in the code repository does not affect functioning of the model wrapper until the updated model wrapper is generated.

17 . A non-transitory computer readable medium having instructions stored thereon that, upon execution by a computing device, cause the computing device to perform operations comprising:
determining a model wrapper for re-deployment, wherein:
the model wrapper has been previously deployed for use, and
the model wrapper comprises a plurality of components, wherein at least one of the plurality of components comprises a data grouping code comprising first instructions for categorizing data analyzed by the model wrapper into one or more of a first plurality of data groups;
receiving or o

Assignees

• Capital One Services Llc

Inventors

• Wu Chen

Classifications

• G06F18/2178

  based on feedback of a supervisor · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06F8/36

  Software reuse · CPC title

• G06F3/0484

  for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range · CPC title

• G06F16/38

  Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually · CPC title