Overnight, shortly before 8:00AM UTC, our internal monitoring system alerted us to issues with our API processing. Upon investigation we realized that API processing, email processing, and several other services (including access to the UI) were being impacted.
Investigation revealed that an unusual pattern of heavy API traffic was the cause. Actions were taken to mitigate the issue, and we were able to restore functionality to our services, and begin processing the backlog. At approximately 1:57PM UTC, the backlog had been fully processed and all API requests and emails began to be processed in real time.
We'll be looking into how to optimize and guard against this particular pattern of API traffic in the future.