Due to a lack of transparency and trust in AI models, telecommunication operators are wary of widespread AI model deployment in their networks, especially when the decisions thereof have both financial and service quality implications. The design of SliceOps solution closely adheres to the ZSM framework. The standardization process for ETSI ZSM is still in its early stages, with preliminary specifications based on a high level of abstraction. The core concept of the closed-loop AI system is to utilize context-aware and metadata-driven policies to more efficiently and quickly identify and incorporate new while updating knowledge and making robust and actionable decisions. SliceOps is a practice toward deploying a more realistic closed-loop scheme to fulfill viable automation solutions for network slicing control. SliceOps manages the lifecycle of AI models in a closed-loop way that includes model performance monitoring, re-training, and delivery. It extends the ZSM framework to manage, besides the service functions, the underlying AI functions, which are natively supported in a standalone slice.