09:25:33:212
[CHECKPOINT]
Assistant speech started

09:25:37:630
[CHECKPOINT]
Assistant speech stopped

09:25:38:665
[CHECKPOINT]
User speech possibly starting

09:25:38:808
[CHECKPOINT]
User speech started

09:25:38:846
[CHECKPOINT]
User speech possibly stopping

09:25:38:891
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:38:891
[LOG]
Transcriber output: - Thank you.

09:25:39:055
[LOG]
Transcriber output: Look,

09:25:39:055
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:39:218
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:39:218
[LOG]
Transcriber output: Look through that.

09:25:39:386
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:39:386
[LOG]
Transcriber output: Look through the details.

09:25:39:555
[LOG]
Transcriber output: Look through the data

09:25:39:555
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:39:727
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

09:25:39:727
[LOG]
Transcriber output: Look through the database,

09:25:39:900
[LOG]
Transcriber output: Look through the database.

09:25:39:900
[LOG]
Endpointing timeout 3000ms (rule: `heuristic`)

Above is part of my logs from a call. This happens every call. I am trying to use a tool which calls my endpoint with info from my database. I see the database get called, but the AI doesn't seem acknowledge that it received anything from it. I'm wondering if these log errors have anything to do with it.

@Pokey It appears that the logs you've provided are related to the transcription and endpointing process during a call, which is part of the voice recognition and processing system. The "endpointing timeout" messages suggest that the system is waiting for a pause in speech to determine the end of a phrase or sentence, which is a normal part of speech recognition. However, the issue you're experiencing with the tool that calls your endpoint and the AI not acknowledging the received information from your database seems to be separate from the transcription process. It could be related to how the tool is integrated with the AI or how the AI is programmed to handle and respond to the data received from the database. To assist you further, I would need more information about the tool you're using and how it's supposed to interact with the AI. If you can provide details about the tool's expected behavior and any error messages you're receiving, I can offer more targeted assistance. If the issue persists and you're unable to provide additional details, I recommend reaching out to support for more in-depth troubleshooting.

@Vapi i think I fixed the tool call. Do you know how I can stop the log error? It is making the chat slower since it needs to wait for the timeout

@Pokey The "endpointing timeout" messages in your logs are not necessarily errors; they indicate that the system is using a heuristic to determine when the user has finished speaking. This is a common feature in voice recognition systems to ensure that the system captures the full intent of the user's speech before processing it. If the endpointing timeout is causing delays and you wish to reduce the waiting time, you can adjust the

silenceTimeoutSeconds

parameter in your configuration. This parameter defines how long the system should wait in silence before considering the user's speech to be complete. To reduce the waiting time, you can set a lower value for

silenceTimeoutSeconds

. However, be cautious with this adjustment, as setting it too low might cause the system to cut off users before they have finished speaking, especially if they naturally take longer pauses in their speech. If you have access to modify the configuration, you can try adjusting the

silenceTimeoutSeconds

to a lower value that still allows users to speak naturally without causing unnecessary delays. If you do not have access or are unsure how to make this change, please reach out to support for further assistance.

@Pokey can you share the call id?