Speech Recognition and Real-Time Transcription Failures in US1

Resolved

Minor impact

Started

Lasted 2h 48m

View official incident ↗

Updates

  1. Resolved

    We have monitored and resolved the issue with speech recognition and real-time transcription features in US1. All services are now operating normally at this time.

  2. Monitoring

    Our engineers have mitigated the issue and are monitoring system performance. Speech recognition and real-time transcription services in US1 are returning to normal, and transcription processing for call recordings may continue to complete more slowly while systems stabilize. We will provide another update in 2 hours or as soon as more information becomes available.

  3. Identified

    We have identified the issue of speech recognition failures for Voice in US1. Customers may experience no speech result when using `<Gather input="speech">`, and real-time transcription sessions may start without the first few seconds of audio transcribed or continue with a different transcription engine. Post-call transcriptions may also complete more slowly than usual. As a workaround, customers may use an alternative speech recognition provider where configurable. We will provide another update in 1 hour or as soon as more information becomes available.

  4. Investigating

    We are investigating an issue with speech recognition and real-time transcription in US1. Users may experience failures with speech-to-text processing affecting <Gather> and real-time transcription. We will provide another update in 1 hour or as soon as more information becomes available.