I was at a tech conference recently, and the local city tech department was showing off the cameras all around the city. They were all tracking individual vehicles with computer vision, across town as the traffic moved from one camera to another.
I didn’t get a chance to ask them whether law enforcement used the data, but I’m pretty sure that’s a given.
Yeah for me the understanding really came when working in a federated GraphQL API. Each team had us own little slice of overall object graph, and overlap / duplication / confusing objects across the whole domain were a lot easier to see in that environment.