This is a bit of a weird one and it's to do with the longevity of livekit being used for live streaming. I have built a few projects now that are powered by livekit streaming, seems great and very nice implementation, my only worry is that live kit seems to be full steam ahead on AI features as opposed to live streaming. Is live streaming still a priority and am I still future proofed using livekit for streaming?

Not weird -- I think it's a pretty fair question to ask, given our focus on multimodal AI. Even today, there are many at-scale livestreaming applications built on top of LiveKit (and using LiveKit Cloud). There are also AI-heavy applications that have a sort of live streaming shape to them like security cameras or robotic teleoperation. We certainly have no intention of making it less easy to build live streaming applications with LiveKit, and there are some features we're working on internally that will be huge unlocks for the live streaming use case. The integration/partnership work we've been doing with AI-generated video/3d avatar services and companies, for example, is something I think will have broad appeal in the live streaming space over time. So, to answer the second part of your question, I do think you are future-proofed using LK for live streaming. On the first part of your question, is live streaming still a priority, in the short-to-medium term, our priority on the commercial side is infrastructure for voice and computer vision-based AI applications. That is to say, most of the new features/products we announce in 2026 will be focused towards serving the needs of those use cases. The positive side of all this is there's a large overlap between the needs of live streaming and other realtime use cases. The negative side is that you won't see us, at least not for a while, building features "up the stack" for some of the tablestakes features one might need for live streaming applications -- integrated payments or advertising, custom players, studio-type of abstractions/components, etc. All these things can be built on top of LiveKit -- it's open source -- but it will be some time, if at all, before we make these types of things core features within the platform. I hope level of visibility is helpful and gives you some clarity on what's the best path forward, for you. Happy to answer any follow-ups!

That does clear some things up and does certainly make me want to continue to use LK for projects, I am currently building out some IoT security cams and using LK in the backend. One other question would be, do you guys ever plan on introducing GPU transcoding? Or is this something that can be done by our selves by changing a few things around on the server?