API cache

Mongodb is oversollicited by queries that are always the same, we should focus on adding a cache in API resolvers for the most common data
+ relevant invalidation logic + relevant cache disable in dev logic
A local in-memory cache should be more than enough but we need to be careful with request cardinality, when people enter requests that are different
Perhaps we should only cache during build time, this can usually be detected using environment variables, vs runtime API calls

Normally this is already done with Redis => check that its ok