System and method for operating an event-driven architecture
There is disclosed a method and system for operating an event-driven architecture. The event-driven architecture comprises a first machine-learning (ML) agent operating a first service and a second ML agent operating a second service. The first ML agent comprises a first model and first model metadata. The second ML agent comprises a second model and second model metadata. The method comprises generating, by the first ML agent, an event associated with event metadata. The event comprises results generated by the first model. The event metadata comprises an event identifier (ID). The first ML agent publishes the event in a virtualized dedicated space. The second ML agent determines whether the event is to be processed by the second ML agent. If a determination is made that the message is to be processed by the second ML agent, the second ML agent processes the event to generate an output.
1 . A computer-implemented method for operating an event-driven architecture, the event-driven architecture comprising a first machine-learning (ML) agent operating a first service and a second ML agent operating a second service, the first ML agent comprising a first model and first model metadata, the second ML agent comprising a second model and second model metadata, the method comprising:
generating, by the first ML agent, an event associated with event metadata, the event comprising results generated by the first model, the event metadata comprising an event identifier (ID);
publishing, by the first ML agent, in a first virtualized dedicated space, the event;
causing the second ML agent to monitor the first virtualized dedicated space;
determining that the event is to be processed by the second ML agent;
processing the event to generate an output, the processing comprising one of further training the second model based on the event or generating a prediction by the second model based on the event;
generating, by the first ML agent, an additional event associated with additional event metadata, the additional event comprising intermediate results generated by the first model;
determining that a level of confidence relating to the intermediate results is lower than a threshold value; and
publishing, based on the level of confidence relating to the intermediate results being lower than the threshold value, the intermediate results in a second virtualized dedicated space.
2 . The method of claim 1 , further comprising:
generating a response based on the output, the response comprising the event ID; and
publishing the response, by the second ML agent, in the first virtualized dedicated space.
3 . The method of claim 2 , further comprising:
determining that the response is to be consumed by the first ML agent based on the event ID;
causing the first ML agent to monitor the first virtualized dedicated space; and
processing the response.
4 . The method of claim 2 , wherein the second ML agent is further configured to execute at least one of caching the event for later usage, augmenting the event or pushing the event back to the first virtualized dedicated space.
5 . The method of claim 1 , wherein determining whether the event is to be processed by the second ML agent is based on a predefined workflow defining a sequence of actions between the first ML agent and the second ML agent.
6 . The method of claim 1 , the second ML agent is one of an inference agent configured to generate a prediction by the second model based on the event, a learning agent configured to execute further training the second model based on the event or an inference/learning agent configured to generate a prediction by the second model based on the event and execute further training the second model based on the event.
7 . The method of claim 1 , wherein the event is a plurality of events, the plurality of events comprising a first set of events associated with a first feed and a second set of events associated with a second feed and wherein the first ML agent monitors the first feed and the second ML agent monitors the second feed.
8 . The method of claim 1 , wherein the event is associated with a ML model and comprises at least a partial representation of the ML model and wherein the metadata is associated with the ML model.
9 . The method of claim 8 , wherein the event associated with the ML model and the metadata associated with the ML model is published in a third virtualized dedicated space (models cloud space).
10 . The method of claim 1 , wherein the first ML agent comprises one or more services mesh which encapsulates state and behavior.
11 . The method of claim 1 , wherein the first ML agent is configured to provide behavioral insights and operational control over a mesh.
12 . The method of claim 1 , wherein the first ML agent and the second ML agent enable one of a network of deployed services with load balancing, service-to-service authentication and monitoring.
13 . The method of claim 1 , wherein publishing the event in the first virtualized dedicated space comprises broadcasting a signal comprising the event.
14 . The method of claim 13 , further comprising enabling a formal representation of knowledge based on a system type information for the event broadcasted within the signal.
15 . A system comprising:
at least one processor, and
memory storing a plurality of executable instructions which, when executed by the at least one processor, cause the system to:
generate, by a first machine-learning (ML) agent operating a first service, an event associated with event metadata, wherein the first ML agent comprises a first model and first model metadata, wherein the event comprises results generated by the first model, an wherein the event metadata comprises an event identifier (ID);
publish by the first ML agent, the event in a first virtualized dedicated space;
cause a second ML agent to monitor the first virtualized dedicated space;
determine that the event is to be processed by the second ML agent;
process the event to generate an output, the processing comprising one of further training the second model based on the event or generating a prediction by the second model based on the event,
generate, by the first ML agent, an additional event associated with additional event metadata, the additional event comprising intermediate results generated by the first model;
determine that a level of confidence relating to the intermediate results is lower than a threshold value; and
publish, based on the level of confidence relating to the intermediate results being lower than the threshold value, the intermediate results in a second virtualized dedicated space.
16 . The system of claim 15 , wherein the second ML agent is one of an inference agent configured to generate a prediction by the second model based on the event, a learning agent configured to execute further training the second model based on the event or an inference/learning agent configured to generate a prediction by the second model based on the event and execute further training the second model based on the event.
17 . The system of claim 15 , wherein the instructions further cause the system to:
generate a response based on the output, the response comprising the event ID; and
publish the response, by the second ML agent, in the first virtualized dedicated space.