Pazau status

Past incidents

1. Degraded performanceVoice38 min

   Elevated call setup latency for Pazau Voice (US)

   1. ResolvedSep 9, 2026 · 15:02 PDT

      Call setup times have been normal for 60 minutes. A carrier-side SIP routing change increased the time to first audio for about 6% of inbound calls. No calls were dropped. We will publish a post-incident review within five business days.

   2. MonitoringSep 9, 2026 · 14:11 PDT

      We moved inbound traffic to a secondary SIP trunk. p95 call setup is back under 1.2 seconds. We're monitoring.

   3. IdentifiedSep 9, 2026 · 13:47 PDT

      The latency is caused by a routing change at one of our upstream carriers in the us-east region. We're failing over to a secondary trunk.

   4. InvestigatingSep 9, 2026 · 13:24 PDT

      We're investigating reports of slow call connection for some Pazau Voice customers in the US region. AI and human-handled calls are affected.

2. MinorIntegrations47 min

   Delayed Salesforce and Zendesk sync

   1. ResolvedAug 18, 2026 · 10:36 PDT

      The sync backlog has cleared, and all queued updates were delivered in order. No data was lost. The cause was an expired OAuth refresh token for a shared integration worker pool, which has been replaced with per-tenant credentials.

   2. IdentifiedAug 18, 2026 · 10:05 PDT

      We identified the cause and are replaying queued sync jobs. Conversations in Pazau are unaffected. Only outbound record updates are delayed.

   3. InvestigatingAug 18, 2026 · 09:49 PDT

      Some customers are seeing delays of up to 40 minutes in case updates syncing to Salesforce and Zendesk. We're investigating.

3. Partial outagePazau Agent (EU)26 min

   Increased Action error rate in the EU region

   1. ResolvedJul 22, 2026 · 08:19 IST

      Action success rates are back to normal. A configuration change to connection pooling for outbound Action calls reduced capacity in eu-west-1. Pazau Agent handed affected conversations to human agents with full context, as designed. The change has been rolled back, and we've added a canary stage for this configuration.

   2. MonitoringJul 22, 2026 · 08:02 IST

      We rolled back the change. Error rates are recovering.

   3. InvestigatingJul 22, 2026 · 07:53 IST

      About 9% of Actions in the EU region are failing with timeouts. Affected conversations are being handed to human agents. We're investigating.

4. MinorAPI21 min

   Webhook delivery delays

   1. ResolvedJul 2, 2026 · 17:28 PDT

      All delayed webhooks have been delivered. Peak delivery delay was 11 minutes. The REST API and inbound requests were not affected.

   2. InvestigatingJul 2, 2026 · 17:07 PDT

      Webhook deliveries are delayed for some customers. API requests are unaffected. We're investigating.