Jarvis

A hands-free AI command centre: a visual agent architecture that feels like a tactical game, controlled with claps, voice and hand gestures.

Type: AI command centre
Status: In active development
Role: Concept, design & development
Stack: Local AI · agents · speech (STT) · gesture recognition

RESULT

3 hands-free control modes — clap, voice and hand gestures — on a local brain.

Talk to me about this Read how it works

Jarvis Nexus — the agent command centre: each agent has its own lane, load and input/output (sample view)

01 THE PROBLEM

A chat window is passive. I wanted an assistant you can actually see working — which agents are active, how they cooperate and where the next action comes from — and that you control hands-free.

02 WHAT IT DOES

Agent command centre: A Nexus Core orchestrates specialised agents (research, briefing, memory, risk) — each with its own lane, load and clear input/output. More mission control than admin panel.
Clap control: 2× short claps open your dashboards, music and editor; 3× claps ask your knowledge base a spoken question. Fully hands-free.
Voice in Flemish: Voice input and spoken answers in the Dutch spoken here.
Hand gestures: A separate version controls the interface with hand gestures via the camera.
Local brain: Runs on local AI, so your conversations and data stay on your own machine.

03 WHAT I LEARNED

Hands-free control forces you to keep intent simple: a clap, a gesture, a sentence. The real challenge wasn't the features but recognising reliably and giving clear feedback straight away.

INTERESTED?

Let's talk.

Tell me what you think — or what you'd want to do with it.

Contact me Next project: Sirena