System and method for real-time modeling inference pipeline
Systems and methods of real-time modeling pipeline inferencing are disclosed. At least one model configured to calculate at least one metric from one or more features is deployed. A model inferencing pipeline configured to extract the one or more features from a customer-specific data pipeline is implemented for the at least one mode. The model inferencing pipeline is generated using a training data set extracted from a cross-customer data pipeline. The at least one metric is calculated using the one or more features extracted from the customer-specific data pipeline.
1. A system, comprising a computing device configured to:
generate a training data set based on cross-customer data from a cross-customer data pipeline;
apply a machine learning process using the training data set to generate one or more model inferencing pipeline templates;
deploy at least one model configured to calculate at least one metric from one or more features;
implement, based on the one or more model inferencing pipeline templates, a model inferencing pipeline configured to extract the one or more features from a customer-specific data pipeline; and
calculate the at least one metric using the one or more features extracted from the customer-specific data pipeline.
2. The system of claim 1 , wherein the computing device is configured to:
implement a shared pipeline configured to extract one or more shared features from the cross-customer data pipeline;
store the one or more shared features in a database; and
provide the one or more shared features to the at least one model, wherein the at least one model calculates the at least one metric using the one or more shared features.
3. The system of claim 1 , wherein the computing device is configured to:
generate the at least one model from a set of model training data extracted from the cross-customer data pipeline; and
provide the at least one model to a model store, wherein the at least one model is deployed from the model store.
4. The system of claim 3 , wherein the computing device is configured to receive at least one query including one or more query parameters, wherein the at least one model is deployed in response to at least one of the one or more query parameters.
5. The system of claim 1 , wherein the at least one model is deployed in response to a predetermined trigger received in the customer-specific data pipeline.
6. The system of claim 1 , wherein the model inferencing pipeline is implemented using a predetermined pipeline environment and the model is deployed using a predetermined model environment, and wherein the computing device is configured to convert an output from the predetermined pipeline environment to an input suitable for the predetermined model environment.
7. The system of claim 1 , wherein the computing device is configured to output the at least one metric to a publication/subscription element.
8. A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by a processor cause a device to perform operations comprising:
generating a training data set based on cross-customer data from a cross-customer data pipeline;
applying a machine learning process using the training data set to generate one or more model inferencing pipeline templates;
deploying at least one model configured to calculate at least one metric from one or more features;
implementing, based on the one or more model inferencing pipeline templates, a model inferencing pipeline configured to extract the one or more features from a customer-specific data pipeline; and
calculating the at least one metric using the one or more features extracted from the customer-specific data pipeline.
9. The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed by the processor cause the device to perform operations comprising:
implementing a shared pipeline configured to extract one or more shared features from the cross-customer data pipeline;
storing the one or more shared features in a database; and
providing the one or more shared features to the at least one model, wherein the at least one model calculates the at least one metric using the one or more shared features.
10. The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed by the processor cause the device to perform operations comprising:
generating the at least one model from a set of model training data extracted from the cross-customer data pipeline; and
providing the at least one model to a model store, wherein the at least one model is deployed from the model store.
11. The non-transitory computer readable medium of claim 10 , wherein the instructions, when executed by the processor cause the device to perform operations comprising receiving at least one query including one or more query parameters, wherein the at least one model is deployed in response to at least one of the one or more query parameters.
12. The non-transitory computer readable medium of claim 8 , wherein the at least one model is deployed in response to a predetermined trigger received in the customer-specific data pipeline.
13. The non-transitory computer readable medium of claim 8 , wherein the model inferencing pipeline is implemented using a predetermined pipeline environment and the model is deployed using a predetermined model environment, and wherein the computing device is configured to convert an output from the predetermined pipeline environment to an input suitable for the predetermined model environment.
14. The non-transitory computer readable medium of claim 8 , wherein the instructions, when executed by the processor cause the device to perform operations comprising outputting the at least one metric to a publication/subscription element.
15. A method, comprising:
generating a training data set based on cross-customer data from a cross-customer data pipeline;
applying a machine learning process using the training data set to generate one or more model inferencing pipeline templates;
deploying at least one model configured to calculate at least one metric from one or more features;
implementing, based on the one or more model inferencing pipeline templates, a model inferencing pipeline configured to extract the one or more features from a customer-specific data pipeline; and
calculating the at least one metric using the one or more features extracted from the customer-specific data pipeline.
16. The method of claim 15 , comprising:
implementing a shared pipeline configured to extract one or more shared features from the cross-customer data pipeline;
storing the one or more shared features in a database; and
providing the one or more shared features to the at least one model, wherein the at least one model calculates the at least one metric using the one or more shared features.
17. The method of claim 15 , comprising:
generating the at least one model from a set of model training data extracted from the cross-customer data pipeline; and
providing the at least one model to a model store, wherein the at least one model is deployed from the model store.
18. The method of claim 17 , comprising receiving at least one query including one or more query parameters, wherein the at least one model is deployed in response to at least one of the one or more query parameters.
19. The method of claim 15 , wherein the at least one model is deployed in response to a predetermined trigger received in the customer-specific data pipeline.
20. The method of claim 15 , wherein the model inferencing pipeline is implemented using a predetermined pipeline environment and the model is deployed using a predetermined model environment, and wherein the computing device is configured to convert an output from the predetermined pipeline environment to an input suitable for the predetermined model environment.