The analytics metaschema specifies a mapping from Datomic entities and attributes to dynamic SQL tables and columns, as shown in the example below:

• Watch the video. 
• Read the Cloud docs.
• Read the On-Prem docs.

Today's releases of Datomic Cloud and Datomic On-Prem include two major new features: tuples and  database predicates.

Tuples are a new compound data type, small vectors as values. You can use tuples to create multi-attribute unique keys on domain entities. You can also use tuples to optimize queries that otherwise would have to join two or more high-population attributes.

• If you declare a composite tuple, Datomic will automatically populate the tuple from existing attributes. 
• You can also define your own homogeneous or heterogeneous tuple types that you populate as you see fit.

• Attribute predicates are declared in schema and constrain the values taken on by a single attribute, in all contexts.  For example, you could limit a test/grade attribute to the range 0-100.
• Entity specs comprise lists of required attributes and/or predicates of the entity and the post-transaction db. You can use entity specs to ensure properties across different attributes of an entity or even across entities.  For example, you might enforce that a game player's score/lowest must be less than or equal to their score/highest.  You must explicitly call for entity specs in transactions in which you want them to apply.

Most Datomic queries return tuples, but sometimes you just want maps.  Today's release of the Datomic Cloud client library adds return maps to Datomic datalog.  For example, the following query uses the new :keys clause to request maps with :artist and :release keys:

[:find ?artist-name ?release-name
 :keys artist release
 :where [?release :release/name ?release-name]
 [?release :release/artists ?artist]
 [?artist :artist/name ?artist-name]]

#{{:artist "George Jones" :release "With Love"}
 {:artist "Shocking Blue" :release "Hello Darkness / Pickin' Tomatoes"} 
 {:artist "Junipher Greene" :release "Friendship"}
 ...}

• Sign up for Datomic Cloud on AWS Marketplace.
• Read the new return maps docs.

1. Get the latest release of Datomic Cloud.
2. Work through the HTTP Direct Tutorial.

Everything gets better when data moves closer to processing. Datomic can extensively cache its immutable data without concern for coordination or invalidation. With the 'valcache' feature in today's Datomic On-Prem release you can use your local SSDs as large, durable per-process caches, on both transactors and peers. Valcache can improve performance, reduce the read load on storage and stay hot across process restarts.

• Get the latest release of Datomic.
• Read the Valcache docs.