We had unexpected ~5min downtime from our proxy layer in US region on 1st March at 07:10 AM UTC. We restored the infrastructure to a working state and the API are being processed again. We have identified the cause for the issue and are working towards a resolution.