Between approximately 7:03 PM on 9 February and 9:04 AM on 10 February US Pacific time, our systems experienced an issue resulting in intermittent elevated levels of 5xx errors on our APIs, particularly named user API. This may have also impacted contact lookup functionality in the go dashboard. Between 7:15 AM and 9:04 AM on 10 February US Pacific time, it significantly impacted our push APIs, and caused some sending delays as that system recovered around 9:00AM. We also saw processing delays in excess of 1 hour on some device registration activities between 8:48 and 9:33 AM on 10 February US Pacific time. These issues impacted customers hosted on our US cloud site. The EU site was not impacted. It should also not have impacted messaging prepared ahead of time through automations or the scheduling system. Our team was able to identify that the issue was due to network communication failures between internal services and our central database, though a root cause analysis is still being performed.