This could take the PULSE_DATA_INGESTION_CONFIG from the environment, in a file, or as a cmd line arg. It would be awesome if it was possible to specify whether to JUST ingest jobs and pushes or to do log parsing as well. This will require Docker.

Another way to reduce the number of steps would be to combine the push and job listeners into one command - perhaps called "pulse_listener"? I don't think we ever need to stop listening to one and not the other (the store worker can be paused instead), and the load is so low that we don't need to run them on separate dynos.