Thanks for the elaborate reply, I really like this conversation. To the points that you disagree:

- For the first one, we can say it might be more a cultural change in terms of how a team is being managed and how much product people gets involved into technical details (data pipelines).

- Regarding the second point, breaking an endpoint for a big company in production can cost a lot of money as well. We all know that mistakes will always happen when writing software. The reason why is because real users, as well as real data, are always unpredictable. Therefore, our goal is not to only to try to avoid them from happening, but also find strategies to mitigate them. The reason I'm saying that is because it doesn't matter how much planning you do upfront since data will come in unexpected ways. Instead we can focus on reduce the time to mitigate the issue.