If you're telling me that if not for the alleged efficiencies gained by using LLMs, your team would be crunching? What that says is that you're failing to address crunch because you're trying to limit workload at the edges instead of systemically preventing it.