AI based service composer advisory for complex public cloud services
An approach is disclosed for training an artificial intelligence (AI) model to recommend services for a complex composite service chain in a cloud computing platform. The training uses metadata from various services provided by different service providers. The method involves inputting a first service into the trained AI model, which then suggests recommended services. These recommendations are displayed to a user, who selects one of the suggested services. The selected service is displayed as part of the service chain. The AI model then processes metadata from the selected service to provide additional service recommendations. This iterative process continues, with the AI model being further trained based on user selections to improve its recommendations.
1 . A method, implemented by a processor coupled to a memory, comprising:
training an artificial intelligence (AI) model to recommend a plurality of services to link to a complex composite service chain forming an application solution in a cloud computing platform, wherein the training inputs metadata corresponding to a plurality of services provided by a plurality of service providers;
ingesting a first service to the trained AI model and responsively receiving a first plurality of recommended services from the trained AI model, wherein the first plurality of recommended services is displayed to a user;
receiving a user selection of one of the first plurality of recommended services;
displaying the selected service as a link in the complex composite service chain;
ingesting, to the trained AI model, metadata corresponding to the selected service and responsively receiving a second plurality of recommended services, wherein the second plurality of recommended services are displayed to the user;
further training the trained AI model using the user selection;
constructing the trained AI model using the plurality of service providers, a plurality of resource types, a plurality of configurations, and a plurality of dependencies;
mapping a frequency of association of data sources and corresponding resource types across the plurality of service providers; and
augmenting a metadata of templates and constructing a complex composite service chained pattern, the complex composite service chained pattern being a packaged definition of the templates representing a complex application, wherein the complex composite service chained pattern is displayable in a graphical user interface.
2 . The method of claim 1 further comprising:
receiving, from the trained AI model, a prerequisite service that corresponds to the selected service; and
displaying the prerequisite service and depicting a visual link between the selected service and the prerequisite service.
3 . The method of claim 1 wherein the training and further training further comprises:
inputting a plurality of consumer metrics, a plurality of rating information, and a plurality of parameter mappings to the trained AI model; and
inputting a plurality of user profile data to the trained AI model.
4 . The method of claim 1 further comprising:
training the trained AI model on a plurality of aspects regarding the plurality of services provided by the plurality of service providers, wherein the plurality of aspects includes security aspects, compliance aspects, recommendation aspects, compatibility aspects, interdependency aspects, deployment aspects, data residency aspects, pre-requisites needed, dynamic pricing aspects, service availability aspects, and validation aspects.
5 . The method of claim 1 further comprising:
discovering the plurality of service providers corresponding to the plurality of services.
6 . The method of claim 1 further comprising:
gathering data that includes a service metadata and a user profile data;
preprocessing and normalizing a plurality of data received from a plurality of service providers of the plurality of services; and
encoding the preprocessed and normalized data into embeddings, wherein the embeddings include attribute embeddings, service embeddings, and user embeddings.
7 . The method of claim 1 further comprising:
assessing a compatibility between the plurality of services;
identifying one or more prerequisite services that correspond to at least one of the plurality of services; and
recommending the first plurality of recommended services, wherein the recommending includes a compatible service recommendation and a personalized recommendation of further services which are compatible with the selected service and can be chained together to form an assured deployable pattern.
8 . An information handling system comprising:
one or more processors;
a memory coupled to at least one of the processors; and
a set of instructions stored in the memory and executed by at least one of the processors to perform actions comprising:
training an artificial intelligence (AI) model to recommend a plurality of services to link to a complex composite service chain forming an application solution in a cloud computing platform, wherein the training inputs metadata corresponding to a plurality of services provided by a plurality of service providers;
ingesting a first service to the trained AI model and responsively receiving a first plurality of recommended services from the trained AI model, wherein the first plurality of recommended services is displayed to a user;
receiving a user selection of one of the first plurality of recommended services;
displaying the selected service as a link in the complex composite service chain;
ingesting, to the trained AI model, metadata corresponding to the selected service and responsively receiving a second plurality of recommended services, wherein the second plurality of recommended services are displayed to the user;
further training the trained AI model using the user selection;
constructing the trained AI model using the plurality of service providers, a plurality of resource types, a plurality of configurations, and a plurality of dependencies;
mapping a frequency of association of data sources and corresponding resource types across the plurality of service providers; and
augmenting a metadata of templates and constructing a complex composite service chained pattern, the complex composite service chained pattern being a packaged definition of the templates representing a complex application, wherein the complex composite service chained pattern is displayable in a graphical user interface.
9 . The information handling system of claim 8 wherein the actions further comprise:
receiving, from the trained AI model, a prerequisite service that corresponds to the selected service; and
displaying the prerequisite service and depicting a visual link between the selected service and the prerequisite service.
10 . The information handling system of claim 8 wherein the training and further training further comprises:
inputting a plurality of consumer metrics, a plurality of rating information, and a plurality of parameter mappings to the trained AI model; and
inputting a plurality of user profile data to the trained AI model.
11 . The information handling system of claim 8 wherein the actions further comprise:
training the trained AI model on a plurality of aspects regarding the plurality of services provided by the plurality of service providers, wherein the plurality of aspects includes security aspects, compliance aspects, recommendation aspects, compatibility aspects, interdependency aspects, deployment aspects, data residency aspects, pre-requisites needed, dynamic pricing aspects, service availability aspects and validation aspects.
12 . The information handling system of claim 8 wherein the actions further comprise:
discovering the plurality of service providers corresponding to the plurality of services.
13 . The information handling system of claim 8 wherein the actions further comprise:
gathering data that includes a service metadata and a user profile data;
preprocessing and normalizing a plurality of data received from a plurality of service providers of the plurality of services; and
encoding the preprocessed and normalized data into embeddings, wherein the embeddings include attribute embeddings, service embeddings, and user embeddings.
14 . The information handling system of claim 8 wherein the actions further comprise:
assessing a compatibility between the plurality of services;
identifying one or more prerequisite services that correspond to at least one of the plurality of services; and
recommending the first plurality of recommended services, wherein the recommending includes a compatible service recommendation and a personalized recommendation of further services which are compatible with the selected service and can be chained together to form an assured deployable pattern.
15 . A computer program product comprising:
a computer readable storage medium comprising a set of computer instructions that, when executed by a processor, are effective to perform actions comprising:
training an artificial intelligence (AI) model to recommend a plurality of services to link to a complex composite service chain forming an application solution in a cloud computing platform, wherein the training inputs metadata corresponding to a plurality of services provided by a plurality of service providers;
ingesting a first service to the trained AI model and responsively receiving a first plurality of recommended services from the trained AI model, wherein the first plurality of recommended services is displayed to a user;
receiving a user selection of one of the first plurality of recommended services;
displaying the selected service as a link in the complex composite service chain;
ingesting, to the trained AI model, metadata corresponding to the selected service and responsively receiving a second plurality of recommended services, wherein the second plurality of recommended services are displayed to the user;
further training the trained AI model using the user selection;
constructing the trained AI model using the plurality of service providers, a plurality of resource types, a plurality of configurations, and a plurality of dependencies;
mapping a frequency of association of data sources and corresponding resource types across the plurality of service providers; and
augmenting a metadata of templates and constructing a complex composite service chained pattern, the complex composite service chained pattern being a packaged definition of the templates representing a complex application, wherein the complex composite service chained pattern is displayable in a graphical user interface.
16 . The computer program product of claim 15 wherein the actions further comprise:
receiving, from the trained AI model, a prerequisite service that corresponds to the selected service; and
displaying the prerequisite service and depicting a visual link between the selected service and the prerequisite service.
17 . The computer program product of claim 15 wherein the training and further training further comprises:
inputting a plurality of consumer metrics, a plurality of rating information, and a plurality of parameter mappings to the trained AI model; and
inputting a plurality of user profile data to the trained AI model.
18 . The computer program product of claim 15 wherein the actions further comprise:
training the trained AI model on a plurality of aspects regarding the plurality of services provided by the plurality of service providers, from across cloud service providers, wherein the plurality of aspects includes security aspects, compliance aspects, recommendation aspects, compatibility aspects, interdependency aspects, deployment aspects, data residency aspects, pre-requisites needed, dynamic pricing aspects, service availability aspects and validation aspects.
19 . The computer program product of claim 15 wherein the actions further comprise:
discovering the plurality of service providers corresponding to the plurality of services.
20 . The computer program product of claim 15 wherein the actions further comprise:
gathering data that includes a service metadata and a user profile data;
preprocessing and normalizing a plurality of data received from the plurality of service providers of the plurality of services;
encoding the preprocessed and normalized data into embeddings, wherein the embeddings include attribute embeddings, service embeddings, and user embeddings;
assessing a compatibility between the plurality of services;
identifying one or more prerequisite services that correspond to at least one of the plurality of services; and
recommending the first plurality of recommended services, wherein the recommending includes a compatible service recommendation and a personalized recommendation of further services which are compatible with the selected service and can be chained together to form an assured deployable pattern.