Switchboard for Voice AI

Build with OpenAI’s new full-duplex Live API, where the model can listen and speak at the same time. Switchboard handles the mobile audio stack—including microphone and speaker I/O, echo cancellation, audio routing, session management, transcripts, and tool calls—through a simple React Native API, with optional on-device speaker isolation for noisy environments.

Build with OpenAI’s turn-based Realtime API while Switchboard handles the complexity around it: WebSocket connectivity, native microphone and speaker I/O, echo cancellation, audio routing, and tool calls. Add on-device VAD, semantic turn detection, and tunable barge-in handling when you need faster, more controllable turn-taking than cloud-side detection alone.

On device speech recognition (ASR / STT) and text-to-speech (TTS) so that you can cut costs and latency while simplifying cloud infra. You only send text to the LLM so don't have to worry about webRTC, sockets, or scaling audio in the cloud.

Build a full voice agent on iOS, running on-device with seamless cloud handoff when you need more intelligence. Switch between local and cloud models without losing conversation context, while keeping audio processing on the iPhone.

On-device preprocessing for speech to speech models (aka S2S or audio models). On device voice activity detection (VAD), echo cancellation, and other audio preprocessing runs locally before connecting to cloud-based speech model (such as OpenAI Realtime API) to optimize performance.

Run OpenAI's Whisper speech recognition (ASR) model entirely on-device for maximum privacy and offline functionality across mobile and desktop platforms (iOS, Android, mac, Windows, Linux).

Run a full STT-LLM-TTS pipeline locally. The STT and LLM components each have optional hand-off (or fallback) to cloud alternatives.

Optimized voice AI components for resource-constrained IoT and embedded systems, including smart speakers, wearables, and edge devices.