We discussed using one DB as our default and initially proposed Postgresql, is this definitely the correct approach taking into consideration having to edit a schema and in turn having to alter SQL tables? We want some more investigation into this to ensure that we make the correct choice as to our default DB.
Scope

• what should our default DB be?
• we could keep PostGresql for metadata, possibly
• consider editing an existing schema here?
• does NoSQL solution simplify this whole solution (editing an existing schema and the NoSQL DB auto adapting to align with the changes)
• if MongoDB is a better solution for the schemas
• can we scope what effort it would be to migrate our code (used for our metadata) to use MongoDB instead of Postgresql
• Could we use Mongo Client command structure for our resolver syntax (rather than Knex). Could this be passed directly to Mongo “as is”, negating the need for V8 VMs and sandboxing (c.f. https://docs.mongodb.com/manual/reference/mongo-shell/ & https://docs.mongodb.com/manual/crud/ ).