I ran into a very similar headache last month while trying to scale our automation routines and API scripts. We kept hitting temporary blocks and unexplained connection drops because too many automated requests were originating from the same corporate IP range. It was incredibly frustrating to debug until we realized the network architecture itself was the bottleneck. To fix it, we ended up routing our script traffic through an isolated external [source](https://proxy-solutions.net/en) to distribute the load across cleaner, dedicated addresses. That immediately cleared up the network restrictions and stabilized our script runs. Definitely pays off to look into your routing setup early on.