The issue was caused by a brief DB hang that has been resolved, and all systems are operating normally. We will continue monitoring to ensure stability.

We're seeing an increased rate in call failures and are actively looking into it.