Yeah, choosing the right AI voice can be surprisingly tricky there are a lot of options and not all of them fit every use case. I’ve worked on a few projects where voice selection made a big difference in how natural the agent felt, and it’s usually less about the platform (like ElevenLabs) and more about tone, latency, and how it handles edge cases. There are a couple subtle things that might be affecting what you’re hearing. What kind of agent are you building? @Laurent