Voice
Talk to your fleet.

Talk to your fleet. OmniConflux is designed voice-first for working with agents across your own machines. Voice is planned.
Voice first, across your own machines
Your fleet is made from your own machines. The design treats voice as a way to work with that fleet, bringing a spoken interface into the agent platform. The design lets you say the goal instead of typing it. In the design, speech is transcribed on your machine and only the text goes to the agent you chose. What happens to that text depends on the agent and model you pick.

The same independent check
Voice is an interface to the work. Every agent task still gets a verifier from a different vendor. A spoken request does not replace review of the result.
Open by design

By design, voice follows the same rule as the rest of the fleet: you choose the agents and the models, local or cloud, from any vendor. The design turns a spoken request into a task for whichever agent you pick, in its own worktree, with the same independent check. See what connects on the integrations page.
Follow the launch details
Supported voice controls and setup instructions will be documented at launch. Start with the fleet and privacy principles, then pre-register for the closed beta.
