Building the software that helps build SpaceX

Nice read.

> The strategy is to avoid getting stuck doing analysis paralysis in a data swamp and instead, capture only the data needed, analyze it, and try again.

This was my take away from this. I worked on an analytics project and analysis paralysis is what I got stuck into. How much data is too much? What we need and what we don't? Those were the bigger concern that writing the code itself.