The Kitewheel data lake extracts the following files in JSON format:
Records all interactions with the identifiers the interactions are created with and context (with object ID)
Records all journey steps with the identifiers the journey steps are created with and context (with object ID)
Identifiers, attributes and list of journey steps and interactions. The created at and last updated at timestamps.
Identifier map of all identifiers types to the currently merged KWID
Each object contains a unique identifier, the set of identifier and identifier type pairs that were used to record the interaction, a timestamp (UTC), the name of the interaction and channel, the engagement score and any context that was provided.
Each object contains a unique identifier, the set of identifier and identifier type pairs that were used to record the journey step, a timestamp (UTC), the name of the Journey step and any context that was provided.
Each profile object contains all of the identifier and identifier type pairs associated with that profile, the attributes given to that profile a list of journey step and interaction object identifiers. [ Do we need these?]
org_<organization_id> contains tables profiles and identifier_map
All projects within an organization will share the same pool of customer profiles (much like they do in Journey Discovery Analytics 2.0).
proj_<project_id> contains tables interactions and journey_steps
Each individual project will be separated into different schemas corresponding to the project_id. Because queries will need to cross schemas, best practice for queries going forward is to use explicit schemas.