My first job was building pipelines in informatica's ETL tool and Pentaho. As far as I can remember, It used to be very visual back in the days. Hadoop and larg

My first job was building pipelines in informatica's ETL tool and Pentaho. As far as I can remember, It used to be very visual back in the days. Hadoop and large scale pipeline came into picture and all of sudden these pipeline were super complex to manage, nobody knew what was happening, and we were riding big data hype. With scripting language, it is easy to test out these pipelines and showcase value before deploying them. Secondly, I think there is some sense of satisfaction or accomplishment when people script a pipeline than visually building them, even if it might more efficient in some cases. So, visual representation is a sexier pitch but it might be a hard sell.