I am having alot of answering macineds and voicemails in my calls. what is the best solution to reduce the time the assistant uses to detects and cuts off the call when it is an aswering machine or voicemail to the barest minimum. This is costing alot as the ai spends alot of time talking to the answering machines or voicemails. https://cdn.discordapp.com/attachments/1383119791942012939/1383119792092741672/image.png?ex=684da22f&is=684c50af&hm=975fc9a97795b6db73539805bcdfdbdf9383b88bd4c11dfb34e3ae7d0097e0d4&

To minimize time spent on answering machines and voicemails, set your assistant's voicemail detection to use the Vapi provider and configure it for rapid detection. Use the following configuration:

json
"voicemailDetection": {
  "provider": "vapi",
  "backoffPlan": {
    "startAtSeconds": 0.5,
    "frequencySeconds": 0.5,
    "maxRetries": 2
  },
  "beepMaxAwaitSeconds": 2
}

This will help your assistant detect voicemails within the first seconds of the call and cut off quickly, reducing wasted time and costs. Adjust these values as needed for your specific call flow. Reference: [Vapi Voicemail Detection Docs](https://docs.vapi.ai/calls/voicemail-detection) Source: - [Vapi Voicemail Detection Docs](https://docs.vapi.ai/calls/voicemail-detection)

It's possible to configure this in a Workflow?

Enable advanced answering machine detection OpenAI models, lower the speech-end and silence thresholds to around 1000–1500 ms, set a max detection window 10–30 s, and immediately hang up or drop voicemail upon detection, then monitor and fine-tune based on call performance data..

alright i would try this

checking if this is resolved/solved for you?

@Kings_big💫 @20 does this thread mean that openai models are actually better at detecting voicemail ? 🤔

Yes