We will now design a mail box system that stores items for customers
pscope: options are O (for yourself), C (for container), T (for tenant) If T is the pscope, requests to this model are always scoped by container and tenant. In this way AmiDb achieves data separation by tenants without additional coding effort on your end. Also read, How AmiDb does partitioning for cassandra and elasticsearch
ska, skb, skc are three search keys. You can have one to a max of three search keys
cola, colb, or any name can be used to define additional fields
Not giving a value for type defaults it to "string"