Skip to main content

Product Essence

Telemus AI is a spatial AI workspace. It gives work a place: notes, task boards, documents, helpers, media, approvals, external-event cards, environment state, and project context can live in a room instead of disappearing into a chat thread or a stack of windows.

The product is best understood as a spatial operating environment for AI-assisted work:

  • Persistent workspace: panels, notes, boards, files, maps, media, and room state are durable across supported sessions.
  • Embodied assistance: the main assistant and helper agents operate with workspace context rather than only a text transcript.
  • Spatial recall: the app can use room placement, project context, visible surfaces, and recent activity to help find, move, and resume work.
  • Governed actions: tool use, plugin calls, file/network operations, restores, and costly operations are classified, previewed, confirmed, and audited where required.
  • Continuity first: snapshots, recovery cues, sync, degraded-mode records, workspace health, and artifact provenance are treated as product features, not maintenance details.

What It Is Not​

Telemus AI is not just a VR chat app, and it is not a chatbot with floating windows. The avatar matters because it gives the workspace an interface, but the workspace itself is the product: the durable room, the artifacts inside it, and the governed actions that can change it.

It is also not an unbounded autonomous agent. High-impact actions should stay visible and reviewable, and external integrations depend on configured accounts, credentials, network access, and platform policy.

The Working Loop​

  1. Create context: open files, write notes, build boards, spawn helpers, add objects, or place media and maps in the room.
  2. Use spatial memory: refer to work by where it is, what project it belongs to, what you looked at recently, or what it is related to.
  3. Delegate bounded work: ask the assistant or helper agents to summarize, organize, compare, inspect, or prepare artifacts.
  4. Review actions: approve risky changes before they affect files, networks, credentials, restores, costly services, or other sensitive surfaces.
  5. Resume later: return through saved workspace state, recovery summaries, branch history, synced peer-device attention, and health proof where supported.

Scope And Maturity​

Feature availability depends on the platform, build, credentials, and runtime configuration. XR paths are supported only on the targets that implement them; Linux, macOS, iOS, and Web remain flat-only or non-XR targets. P2P quick sessions are private and temporary; do not treat them as durable shared workspaces or assume they are available on every network.

For a feature-by-feature overview, continue to the Spatial Workspace, AI and Helpers, Continuity and Recovery, and Trust and Safety pages.