Webhook 5xx retry logic and dead letter queue implementation

We are hitting a wall with webhook delivery failures. The CXone platform sends events to our internal endpoint, but when the target service times out, it returns a 502 Bad Gateway. The platform retries a few times, but eventually gives up and drops the event. We need a way to capture these failed payloads for manual processing or a secondary retry mechanism.

I’ve checked the webhook settings in the portal, but there’s no obvious toggle for a dead letter queue or a secondary failure endpoint. I’m looking at the API docs for PUT /api/v2/analytics/webhooks but I don’t see a field for failureEndpoint or similar.

Here is the current payload structure we receive:

{
 "event": "interaction.created",
 "data": {
 "id": "abc-123",
 "type": "voice"
 }
}

When the 502 happens, we get nothing. Is there a way to configure a fallback endpoint via the API that gets called only after all retries are exhausted? Or do we need to handle this entirely on the receiving end by returning a 200 OK immediately and processing asynchronously? The async route adds latency to the agent UI, so we’d prefer a platform-side solution if one exists.

Any experience with setting up a secondary webhook for failures?