Immersive Voice AI Homepage – LiveKit
Current LiveKit homepage voice experience with interactive switching between AI personas via voice using function calling.
Overview
I designed and built the current LiveKit homepage voice experience, creating an interactive front-end that puts our technology front and center. The concept was to make it effortless for visitors to explore different use cases — for example, using function calling to switch between AI personas entirely by voice.
I paired with a backend engineer to help build the agent functionality in Python, and owned all design and front-end implementation.
To highlight the flexibility of our platform, the interface dynamically updates to display the underlying model providers (LLMs, STT, TTS) as users switch personas, alongside real-time latency metrics. This both demonstrates the breadth of our integrations and reinforces the speed and responsiveness of LiveKit's real-time capabilities.