This is not really an issue, but I think it might be useful to register here and be reachable by other people who may have the same question.
In the context of the proposed template separation on application, models, and data, how do you think we should handle a case where an application has a component that is an AI agent that is almost, but not quite, an application itself? For example, the Agent part may include some traditional ML models to support the agent’s decisions, specific datasets used to train those models, prompts versioning, and its own degree of autonomy with different decision trees or workflows. Should we treat it as just another model, or perhaps as a second application?
Thanks and congratulations on the excellent work on these templates. They are really useful!
This is not really an issue, but I think it might be useful to register here and be reachable by other people who may have the same question.
In the context of the proposed template separation on application, models, and data, how do you think we should handle a case where an application has a component that is an AI agent that is almost, but not quite, an application itself? For example, the Agent part may include some traditional ML models to support the agent’s decisions, specific datasets used to train those models, prompts versioning, and its own degree of autonomy with different decision trees or workflows. Should we treat it as just another model, or perhaps as a second application?
Thanks and congratulations on the excellent work on these templates. They are really useful!