Configurable deployment of data science models
An example computing platform is configured to (a) cause a client device associated with a user to display an interface for deploying a new data science model, where the interface presents the user with a list of deployment templates, and where each of the deployment templates includes data specifying (i) a respective executable model package and (ii) a respective set of execution instructions for the respective executable model package, (b) receive, from the client device, data indicating (i) a user selection of a given deployment template for use in deploying the new data science model and (ii) a given set of configuration parameters for use in deploying the new data science model, and (c) use the given executable model package, the given set of execution instructions, and the given set of configuration parameters to deploy the new data science model.
1. A client device comprising:
one or more communication interfaces;
one or more processors;
data storage comprising one or more non-transitory computer-readable mediums; and
program instructions stored in the data storage that, when executed by the one or more processors, cause the client device to:
display an interface for requesting that a computing platform deploy a data science model, wherein the interface presents a list of deployment templates, and wherein each of the listed deployment templates is stored at the computing platform and comprises data specifying (i) a respective executable model package and (ii) a respective set of execution instructions for the respective executable model package;
receive, via the interface, a user request to deploy a new data science model, wherein the user request comprises:
a user selection, from the list of deployment templates, of a given deployment template for use in deploying the new data science model, wherein the given deployment template is stored at the computing platform and comprises data specifying (i) a given executable model package and (ii) a given set of execution instructions for the given executable model package; and
user input specifying a given set of configuration parameters for use in deploying the new data science model; and
after receiving the user request to deploy the new data science model, cause the computing platform to deploy the new data science model based on the given deployment template and the given set of configuration parameters by transmitting, to the computing platform over a network-based communication path, an indication of the user request to deploy the new data science model that includes (i) an identification of the given deployment template and (ii) data indicating the given set of configuration parameters.
2. The client device of claim 1 , wherein the given set of configuration parameters for use in deploying the new data science model comprises either or both of (i) an identification of an input dataset for the new data science model or (ii) an identification of an output storage location for the new data science model.
3. The client device of claim 1 , wherein the given set of configuration parameters for use in deploying the new data science model comprises a computing resource allocation to use for deploying the new data science model.
4. The client device of claim 1 , wherein the given set of configuration parameters for use in deploying the new data science model comprises an indication of whether the new data science model is to be executed (i) manually or (ii) according to a schedule.
5. The client device of claim 1 , wherein the list of deployment templates that is presented to the user is defined based on user permissions assigned to the user.
6. The client device of claim 1 , wherein, after the user selection of the given deployment template is received, the interface presents an initial set of configuration parameters corresponding to the given deployment template, and wherein the user input specifying the given set of configuration parameters comprises one or both of (i) user input that adopts at least one configuration parameter in the initial set or (ii) user input that modifies at least one configuration parameter in the initial set.
7. The client device of claim 1 , further comprising program instructions stored in the data storage that, when executed by the one or more processors, cause the client device to:
present an indication of a status of the new data science model after it has been deployed.
8. The client device of claim 7 , wherein the status of the new data science model is expressed in terms of one or both of (i) execution status information or (ii) performance status information.
9. The client device of claim 1 , wherein:
the given executable model package of the given deployment template comprises (a) a trained model object and (b) a pre-encoded set of pre-processing operations that are available for use with the trained model object; and
the given set of execution instructions of the given deployment template comprises instructions for (a) which one or more pre-processing operations from the pre-encoded set of pre-processing operations are to be used with the trained model object and (b) how the one or more pre-processing operations are to be arranged.
10. A non-transitory computer-readable medium, wherein the non-transitory computer-readable medium is provisioned with program instructions that, when executed by at least one processor, cause a client device to:
display an interface for requesting that a computing platform deploy a data science model, wherein the interface presents a list of deployment templates, and wherein each of the listed deployment templates is stored at the computing platform and comprises data specifying (i) a respective executable model package and (ii) a respective set of execution instructions for the respective executable model package;
receive, via the interface, a user request to deploy a new data science model, wherein the user request comprises:
a user selection, from the list of deployment templates, of a given deployment template for use in deploying the new data science model, wherein the given deployment template is stored at the computing platform and comprises data specifying (i) a given executable model package and (ii) a given set of execution instructions for the given executable model package; and
user input specifying a given set of configuration parameters for use in deploying the new data science model; and
after receiving the user request to deploy the new data science model, cause the computing platform to deploy the new data science model based on the given deployment template and the given set of configuration parameters by transmitting, to the computing platform over a network-based communication path, an indication of the user request to deploy the new data science model that includes (i) an identification of the given deployment template and (ii) data indicating the given set of configuration parameters.
11. The non-transitory computer-readable medium of claim 10 , wherein the given set of configuration parameters for use in deploying the new data science model comprises one or more of (i) an identification of an input dataset for the new data science mode, (ii) an identification of an output storage location for the new data science model, (iii) a computing resource allocation to use for deploying the new data science model, or (iv) an indication of whether the new data science model is to be executed (a) manually or (b) according to a schedule.
12. The non-transitory computer-readable medium of claim 10 , wherein the list of deployment templates that is presented to the user is defined based on user permissions assigned to the user.
13. The non-transitory computer-readable medium of claim 10 , wherein, after the user selection of the given deployment template is received, the interface presents an initial set of configuration parameters corresponding to the given deployment template, and wherein the user input specifying the given set of configuration parameters comprises one or both of (i) user input that adopts at least one configuration parameter in the initial set or (ii) user input that modifies at least one configuration parameter in the initial set.
14. The non-transitory computer-readable medium of claim 10 , wherein the non-transitory computer-readable medium is further provisioned with program instructions that, when executed by at least one processor, cause the client device to:
present an indication of a status of the new data science model after it has been deployed.
15. The non-transitory computer-readable medium of claim 10 , wherein:
the given executable model package of the given deployment template comprises (a) a trained model object and (b) a pre-encoded set of pre-processing operations that are available for use with the trained model object; and
the given set of execution instructions of the given deployment template comprises instructions for (a) which one or more pre-processing operations from the pre-encoded set of pre-processing operations are to be used with the trained model object and (b) how the one or more pre-processing operations are to be arranged.
16. A method carried out by a client device, the method comprising:
displaying an interface for requesting that a computing platform deploy a data science model, wherein the interface presents a list of deployment templates, and wherein each of the listed deployment templates is stored at the computing platform and comprises data specifying (i) a respective executable model package and (ii) a respective set of execution instructions for the respective executable model package;
receiving, via the interface, a user request to deploy a new data science model, wherein the user request comprises:
a user selection, from the list of deployment templates, of a given deployment template for use in deploying the new data science model, wherein the given deployment template is stored at the computing platform and comprises data specifying (i) a given executable model package and (ii) a given set of execution instructions for the given executable model package; and
user input specifying a given set of configuration parameters for use in deploying the new data science model; and
after receiving the user request to deploy the new data science model, causing the computing platform to deploy the new data science model based on the given deployment template and the given set of configuration parameters by transmitting, to the computing platform over a network-based communication path, an indication of the user request to deploy the new data science model that includes (i) an identification of the given deployment template and (ii) data indicating the given set of configuration parameters.
17. The method of claim 16 , wherein the given set of configuration parameters for use in deploying the new data science model comprises one or more of (i) an identification of an input dataset for the new data science mode, (ii) an identification of an output storage location for the new data science model, (iii) a computing resource allocation to use for deploying the new data science model, or (iv) an indication of whether the new data science model is to be executed (a) manually or (b) according to a schedule.
18. The method of claim 16 , wherein, after the user selection of the given deployment template is received, the interface presents an initial set of configuration parameters corresponding to the given deployment template, and wherein the user input specifying the given set of configuration parameters comprises one or both of (i) user input that adopts at least one configuration parameter in the initial set or (ii) user input that modifies at least one configuration parameter in the initial set.
19. The method of claim 16 , further comprising:
presenting an indication of a status of the new data science model after it has been deployed.
20. The method of claim 16 , wherein:
the given executable model package of the given deployment template comprises (a) a trained model object and (b) a pre-encoded set of pre-processing operations that are available for use with the trained model object; and
the given set of execution instructions of the given deployment template comprises instructions for (a) which one or more pre-processing operations from the pre-encoded set of pre-processing operations are to be used with the trained model object and (b) how the one or more pre-processing operations are to be arranged.