There was a service degradation on our caching layer that affected the latency of our APIs. The caching layer is back to normal and latency dropped to normal levels.