Object model
All audio analytics consist of listening sessions for a radio stream or podcast episode.
To enable ownership and grouping, the API also recognise these entities:
-
Podcast
channel(the "show") - a group under which episodes are published. All episodes must belong to at least one channel. -
Top-level
providerthat allstreamandchannelobjects belong to. Typically this grouping is the customer/publisher account under which content is published.
graph TD
A1(Provider) --> A2(Stream)
A1(Provider) --> A3(Podcast channel) --> A4(Podcast episode)
Partitioning
The database stores stream and podcast data in discrete datasets. A single API query can only be over one of these products.
Note
Metadata for objects are not managed by the API, but retrieved from the content platform itself via webhooks.