We are seeing repeated mid-call drops and Twilio errors that correlate directly with Vapi backend timeouts. Twilio errors observed: 31901 – WebSocket connection timeout 15003 – HTTP 502 from Vapi /status callback 31951 – Invalid stream SID (post-timeout) Twilio explicitly reports 30s timeouts against: phone-call-websocket.aws-us-west-2-backend-production1.vapi.ai Vapi call logs confirm: Calls are accepted and start normally Assistant speaks initially Mid-call termination and partial/empty transcripts One call never fully initialized (empty transcript, cost = 0) Affected Vapi Call IDs: 019c8b68-5492-755e-be3d-d30e31eae135 019c8b6b-1244-7aad-b738-a0911e5c26d0 019c8b6b-ef19-7009-bc43-e994d3a81e1a 019c8b6d-d243-7112-9e50-39fd2148ecd1 This appears to be a Vapi backend / streaming infrastructure issue (AWS us‑west‑2), not TwiML or customer configuration. Please confirm incident scope, affected regions, and whether mitigation or failover is in progress.

In addition, we are facing a serious transparency issue.... There is no proactive communication from the team when a failure occurs, no clear follow-up, and response times during major outages are extremely slow. At the moment, we have zero visibility into the current issue, which is very concerning. The platform still shows 100% uptime, the outage reporting channel is not working, and the tools that are supposed to provide real-time transparency and status updates do not appear to be functioning properly. This makes incident management on our side extremely difficult and directly impacts our operations.

Yep, there are many issues. Especially for users like me who are running full SaaS on Vapi's infrastructure. When I have 100 clients who can't recieve calls that is a big issue and completely unacceptable.

Happy to offer better experience and 24*7 customer support. We have been building Omnichannel Infra which is >> VAPI + n8n at unique platform. see https://cdn.discordapp.com/attachments/1475543488870354995/1475632679738413259/image.png?ex=699e3186&is=699ce006&hm=f903bf5acb4944114ad1cd96da978ac7aad6962812e44e9b28b960f2244e7883&

Hi, Sorry for the inconvenience you experienced. The issue has now been resolved. You can always find details, if any outage here: https://status.vapi.ai/incident/831494 We also recommend subscribing to the Weekly status updates channel.