Seyra

Ask a chatbot about something in front of you and it answers in text, leaving you to guess which part it means. Seyra is a voice-controlled mixed-reality assistant for Meta Quest 3 that answers in your room instead: look at an object, ask about it out loud, and labels appear on the relevant objects or parts. You can also ask Seyra to build a 3D model on your table, put a checklist on your wall, or show you how your room might look with a different paint color. Follow-up requests let you refine what it creates through the same conversation. Seyra uses Muse Spark to interpret requests and plan actions, SAM 3.1 to locate objects and parts in camera images, and Muse Image to generate design previews. Instead of a reply you have to interpret, you get something you can point to, interact with, or come back to as you work. I built it in Unity with Meta’s XR SDKs, so it can extend to other devices that support the features it relies on.