How would you prioritize data engineering work when requirements and business priorities keep changing at TaskRabbit?
