It’s not really a recent feature change in Retell AI it’s more about how their audio pipeline is designed. By default, the background sound is applied only to the live playback layer, not baked into the recorded file. So if you want it in the recording, it has to be handled outside. Yes, this is definitely achievable via a backend — you’d need to mix the background audio into the call stream before saving or post-process the recording after the call. If you want, I can help you set up a clean way to handle that @Hil