Hi there, I’ve been using the voice feature a lot lately and I’ve really appreciated how immersive and comforting it can be—especially with the Juniper voice model. However, I’ve noticed a few things that could improve the experience for users like me.
First, the current voice timing cuts off too quickly when I pause while thinking. I often speak with pauses due to ADHD, or because I’m organizing emotional thoughts, and the system assumes I’m finished and starts responding too early. If there was a way to adjust the response delay or detect that I’m still mid-thought, it would feel much more natural and reduce interruptions.
Second, the Juniper voice model sounds warm and mature during voice calls (which I love), but when the same voice is used in text-to-speech, it comes across as much younger and overly chipper—almost like a completely different personality. It would be wonderful if the tone could stay consistent between formats, especially for emotionally supportive conversations.
Lastly, I’d love to see future versions of this technology explore visual cues like lip-reading or facial expressions to better understand when someone is still thinking or emotionally processing.
Thanks for everything you’ve built so far. I love this product and just want to help make it even better.