Not specific to Twitter but as I understand it. Generally a data engineering team will setup the pipelines and ETL processes to turn transactional data into aggregate data necessary to run analytics.