US 12,282,832 B1
System and methods for model management
Nicholas Elprin, San Francisco, CA (US); Christopher Yang, Cuptertino, CA (US); and Madhavan Thirumalai, San Francisco, CA (US)
Assigned to Domino Data Lab, Inc., San Francisco, CA (US)
Filed by Domino Data Lab, Inc., San Francisco, CA (US)
Filed on Aug. 14, 2023, as Appl. No. 18/449,369.
Application 18/449,369 is a continuation of application No. 17/702,185, filed on Mar. 23, 2022, granted, now 11,763,201.
Application 17/702,185 is a continuation of application No. 16/529,518, filed on Aug. 1, 2019, granted, now 11,315,039, issued on Apr. 26, 2022.
Claims priority of provisional application 62/714,580, filed on Aug. 3, 2018.
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 20/00 (2019.01); G06F 8/60 (2018.01); G06F 3/04847 (2022.01)
CPC G06N 20/00 (2019.01) [G06F 8/60 (2013.01); G06F 3/04847 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A system for automated computing environment management, the system comprising:
one or more computer processors that are programmed to provide an application comprising:
(a) a first software module for capturing multiple components related to a training and deployment process of a plurality of machine learning models executed within a computing environment, wherein the multiple components are selected from a group consisting of: one or more models used in an experiment, configuration the experiment were run on, input dataset to the one or more models, packages and code were used to execute the experiment, time duration of execution of the code, and output files generated by the execution of the experiment;
(b) a second software module for storing, in a registry, a version of the computing environment upon detection of an update of at least one of the multiple components;
an electronic display with a graphical user interface for setting up a rule to detect a model performance drift or a model data drift associated with a machine learning model selected from the plurality of machine learning models.