livekit

15 live roles

You'll set the vision and roadmap for LiveKit Inference, from which models and providers we offer to how we think about the space evolving strategically over the medium and long term. You’ll partner directly with the engineering team that builds it. This product involves both commercial relationships with frontier model labs as well as direct model hosting and optimization: we recently launched a voice-optimized deployment of Gemma 4 31B, and we're scaling up to become the go-to inference provider for voice workloads. The goal is to make LiveKit Inference the obvious default over wiring up model providers directly, easier, faster, cheaper, or better, and eventually good enough that companies come to LiveKit for voice inference itself. The model landscape moves faster than almost anything in software, and you'll decide how LiveKit stays ahead of it. You'll Thrive Here If You: are obsessed with making the best product for our customers