Handling Shift in Data Distribution with DataOps/MLOps

What are the available techniques and tools we can use to manage the shift in the distribution of dataset over time and also handling sub-variations of dataset provided sub-variations are not annotated