How did you approach the design of your previous data pipelines, from requirements to production operations?
