Talk to Your Cameras. Literally.

Rhombus + Claude Series

Imagine walking through your facility and asking, out loud, “are there any active alerts at the warehouse right now?” — and getting an answer. No laptop. No dashboard. No app. Just a question and a response, the same way you’d ask a coworker.

That’s not a future product roadmap. It’s a feature shipping today, because when your camera system is API-first and your AI agent can use voice, the interface disappears. You don’t interact with the software anymore. You just ask the question you actually wanted to ask.

For security teams, this changes what an operations center looks like. Walking the floor, driving between sites, sitting in a meeting — you can query your entire deployment hands-free. For everyone else in the organization, it removes the biggest barrier to using security data at all: the fact that most people never log into the VMS because they don’t know how.

For the developers:

rhombus-cli ships with a built-in voice mode. One command launches it:

rhombus voice

It records your question, transcribes it locally via Whisper (audio never leaves your machine), passes it to Claude with full access to your deployment, and speaks the answer back.

The payoff: The best interface for most security questions is no interface at all. When your camera system can be queried by voice, it becomes usable by the 95% of your organization who will never open a dashboard. That’s not a feature — it’s a different product category.