Google Introduces New Services To Simplify MLOps

The company announced a fully managed service for ML pipelines, a Continuous Monitoring service, and a new ML Metadata Management service.

Recently, Google introduced a new set of services that will simplify Machine Learning Operations - MLOps.
 
 
 Source: Google
 
The company announced that a fully managed service for ML pipelines will be available in preview by October this year which will enable customers to build ML pipelines using TensorFlow Extended (TFX’s) pre-built components and templates that significantly reduce the effort required to deploy models.
 
Another new offering is a Continuous Monitoring service that will monitor model performance in production. It will let you know if it is going stale, or if there are any outliers, skews, or concept drifts. This will enable teams to quickly intervene, debug, or retrain a new model.
 
Google said that its new ML Metadata Management service in AI Platform enables AI teams track all the important artifacts and experiments they run, providing a curated ledger of actions and detailed model lineage. The service enables you to determine model provenance for any model trained on AI Platform for debugging, audit, or collaboration.
 
AI Platform Pipelines automatically tracks artifacts and lineage and you can also use the ML Metadata service directly for custom workloads, artifact and metadata tracking. You can expect the ML Metadata service to be available in preview by the end of September.
 
The company is also bringing a Feature Store in the AI Platform which will serve as a centralized, org-wide repository of historical and latest feature values, thereby enabling reuse within ML teams. Feature Store is expected by the end of this year. You will also be provided tooling to mitigate common causes of inconsistency between the features used for training and prediction.