Labels: feature-request · UX · Wave 3
Related existing issues: #2136 (OPEN — French-language ask for STT/voice transcription, narrower subset of this)
Summary
Add a voice mode for talking to the CEO / Command Center hands-free — dictate ideas on the go, get spoken status back, without opening a keyboard.
Problem
Interaction with the CEO today is text-only. There's no way to capture a fast-moving idea, or get an oversight check, without sitting down and typing — which kills the "single point of contact" convenience (FR-06) exactly when it matters most: away from the desk.
Proposed capability
- A voice mode for the CEO conversation: speak to dictate an idea, a correction, or a quick approval; hear a spoken status back instead of reading it.
- Exact interaction shape is intentionally left open for design — push-to-talk, a standing spoken-briefing mode, and voice dictation straight into ideation are all reasonable entry points, not a committed spec.
- Should ride the same single conversational surface as the rest of the CEO/chat experience (FR-06, FR-22) rather than becoming a separate app.
Prior-art (inspiration)
- openai/whisper — the accuracy bar for turning spoken dictation into clean text, including mixed-language and noisy input.
- livekit/agents — real-time, low-latency voice-agent pipeline (STT → LLM → TTS) built for natural back-and-forth conversation rather than a stop-and-wait recording.
- OpenInterpreter/01 — push-to-talk as the primary interaction model for an always-available assistant, worth borrowing as one selectable input mode.
Why now
Voice is a low-effort way to widen when and how the user can steer the org — capturing an idea or checking status in a spare moment, not just at a desk. Lower priority than the core control-plane and trust work; the shape can stay loose until there's a concrete use case pulling it into focus.
Labels: feature-request · UX · Wave 3
Related existing issues: #2136 (OPEN — French-language ask for STT/voice transcription, narrower subset of this)
Summary
Add a voice mode for talking to the CEO / Command Center hands-free — dictate ideas on the go, get spoken status back, without opening a keyboard.
Problem
Interaction with the CEO today is text-only. There's no way to capture a fast-moving idea, or get an oversight check, without sitting down and typing — which kills the "single point of contact" convenience (FR-06) exactly when it matters most: away from the desk.
Proposed capability
Prior-art (inspiration)
Why now
Voice is a low-effort way to widen when and how the user can steer the org — capturing an idea or checking status in a spare moment, not just at a desk. Lower priority than the core control-plane and trust work; the shape can stay loose until there's a concrete use case pulling it into focus.