Building a Data Platform with Datomic
summary
Concrete pattern for a Datomic-based platform: use Datomic’s EAVT schema as the primitive and layer a self-describing meta-model (a DT attribute plus fields, parent, list, component) so entities carry their own type hints; implement schema evolution by attaching migration functions as Datomic database functions to compute defaults or rewrite instances; enforce constraints at transact time and record actor/context via reified transactions for auditing; show workarounds for variant/any types and typed references; used successfully to support highly dynamic, runtime-configured data models.
AI-written summary · about
chapters (21)
- 0:00Intro
- 0:35WHAT IS A DATA PLATFORM
- 1:39BACKGROUND
- 3:20WHERE DOES APPIAN POWER COME FROM?
- 5:34ROADMAP
- 6:23DATA MODELS
- 9:05SCHEMAS
- 11:11THE KEY IS ADAPTABILITY
- 12:36BASIC API
- 14:23SHORT-HAND NOTATION
- 14:49ATTRIBUTES ARE FACTS!
- 15:20FREE ASSOCIATIONS
- 16:47CORE DATA MODEL: DATATYPE
- 21:26DATATYPE DATATYPE
- 23:00META-MODEL
- 24:04DATATYPE HIERARCHIES
- 30:15TYPED REFERENCES
- 33:31DATABASE FUNCTIONS
- 35:59DECIDE WHEN TO MIGRATE
- 37:42ENUMS
- 39:11AUDIT TRAIL?
