basic app runs on meta quest pro, but issue with agent and voice interaction

Hi - what a cool project, congratulations! Really love it. I do have a Meta quest pro and tried to get this to run, but only succeeded halfway: I was able to open it in unity, add a gemini key (which I know works because I am able to send a source prompt to tell me a joke, and also to use the text to speech tester from unity), and then compile it and transfer the apk to my headset. It runs (after some hiccups with windowed mode now in fullscreen) and I'm able to see how the eye-tracking (if selected) picks up on certain lines as context. However, what I was not able to get to run is the voice-based interaction (talking about the document/context I am lookking at). And I didn't see any flying "bard" figure (like in the presentation video). I was wondering if this could be due to my system (working on a macbook pro, which gave me some pain for compiling), or whether there are additional steps (beyond just adding the gemini key and compiling) (perhaps also Meta / Samsung?). Again - really awesome work - it's really happening ;-) Thank you, RS