The data provided by the publisher sources that have been integrated with Cortex can be processed by worker tasks to create a new source of transformed data. A worker task will in reaction to a subscription event produce a list of dependencies; a specification of the data needed for computing this task. Cortex will then execute the subscriptions required to this data and collect the complete state of all dependencies before providing it to the worked task for computation. After the computation is completed the worked task returns the result to Cortex for distribution to all subscribers.
