Hey Vapi Team, we're stuck and really need some urgent guidance on our concurrency issues. We're running into a very high volume of 429 - Rate Limit Exceeded errors on our

POST /call

requests, and it's making the system unstable under load. We've already set our max Vapi concurrency limit to just 20 and lowered our AWS Lambda concurrency, but the 429s are persisting. We even implemented the recommended exponential backoff logic for retries, but we’re still fighting the instability. For context, our complexity is higher because we have essential secondary call logic (

_second_call

) that triggers a backup call using a landline number when the primary mobile call fails. This is crucial since some mobile operators reject calls coming from our original mobile numbers. Can you advise on why we are consistently hitting the 429 limit even when capped at such a low concurrency of 20? We need help understanding the true bottleneck so we can stabilize our service. Thanks!