Is Synthflow AI Down?

**What happened** An internal configuration change restricted which AI models the Synthflow platform was permitted to call. The restriction was applied more broadly than intended and blocked the models used by our post-call processing pipeline. As a result, AI-generated post-call outputs failed to produce calls completed during the window. The configuration was reverted at 11:10 UTC and all functionality returned to normal immediately. **What was affected** Calls placed and received during this window connected, ran, and completed normally. The failures were confined to AI processing within the post call webhook that runs **after** a call ends. The following were affected: _Info extractors_: Extracted variables not populated _Call summary_: Not generated _LLM-as-a-judge_: Did not run _Custom evaluations_: Did not run _Post-call SMS actions_: Not triggered or sent If you consume the post-call webhook, the webhook is still fired and delivered but these fields were empty or missing in the payload for affected calls. **What was not affected** • Inbound and outbound call connectivity • Live conversation with your agents during the call • Call recordings and audio • Transcripts • Post-call webhook delivery itself (the webhook fired as normal) • Call logs, billing records, and the dashboard **What you should do** If your workflows depend on extracted variables, summaries, or post-call SMS for calls completed between 08:20 and 11:10 UTC on 7 August, treat those records as incomplete. Recordings and transcripts for those calls are intact. **What we're changing** We've completed an internal review of this incident and are making improvements in three areas: • **Detection** — strengthening monitoring so issues of this type are surfaced and acted on faster. • **Change controls** — tightening the review and approval process for platform configuration changes. • **Impact review** — improving how we assess the downstream reach of a change before it is applied.

Resolved. --- [2026-07-15T09:39:11Z] (identified) **Summary** Between 12:29 UTC and 14:00 UTC, a subset of customers experienced agents that were unresponsive or significantly delayed. The disruption was caused by an upstream incident at Microsoft Azure affecting the Azure OpenAI service in the East US 2 region, which caused authentication failures on requests to certain GPT-4.1 model deployments. **What happened** At approximately 12:29 UTC, a portion of our model traffic to Azure OpenAI's East US 2 region began failing with authentication errors. Our credentials had not changed, and we confirmed our configuration was correct on our side — ruling out any issue within our platform. Further investigation confirmed an active Azure-side incident impacting OpenAI services in East US 2. **Impact** • **Duration:** ~1 hour 31 minutes (12:29 UTC – 14:00 UTC) • **Scope:** A subset of agent traffic dependent on the affected Azure region • **Symptoms:** Agents not responding, elevated response latency, request timeouts **Resolution** The incident resolved automatically once Microsoft restored service in the East US 2 region. No changes were required on our side, confirming the issue originated upstream at Azure. **Follow-up actions** While the root cause was external, we are taking steps to reduce the impact of similar upstream provider incidents in the future: 1. Expanding automatic failover coverage so authentication issues from any single provider region are transparently rerouted to healthy regions. 2. Broadening the set of error conditions that trigger immediate failover. 3. Improving monitoring of direct provider incident notifications, as public provider status dashboards did not reflect the outage in a timely manner. We apologize for the disruption and appreciate your patience while Azure restored service.

**Service Incident Report — US Voice Service Disruption** **Date:** July 6, 2026 **Status:** Resolved **Summary** On July 6, 2026, our US voice service experienced a brief disruption lasting approximately 18 minutes. During this window, calls to and from US-based voice agents failed to connect. The issue has been fully resolved, and all services have returned to normal operation. **What happened** During planned maintenance on our US infrastructure, a routine cleanup step inadvertently removed a network component that our US telephony service depended on. This severed the connection our call-routing system needed to handle calls, causing them to fail to connect. **Impact** • **Duration:** Approximately 18 minutes • **Region affected:** US only. Our EU and global regions were not affected. • **Effect:** Inbound and outbound calls to US voice agents did not connect during the affected window. All EU and global traffic was unaffected. **Resolution** Our engineering team identified the cause quickly and restored the connection, after which calls recovered immediately. **Preventing recurrence** We take reliability seriously and have already begun work to prevent this from happening again, including: • Reviewing our infrastructure configuration to remove reliance on any undocumented dependencies • Introducing additional review and safeguards for changes to production systems • Adopting a stricter decommissioning checklist that verifies real usage before any shared infrastructure is removed We apologize for any disruption this may have caused and appreciate your understanding. If you have any questions or noticed continued issues, please reach out to our support team. --- [2026-07-10T12:24:49Z] (resolved) The issue is now resolved, traffic is flowing again. We will provide an RCA shortly. --- [2026-07-10T12:24:49Z] (identified) We are experiencing degraded performance on our calling infrastructure in the US cluster. This only affects clusters on the US cluster.