Ambient Interface Infrastructure

AI2X — The Display Layer for AI Agents

Turn any screen into an AI canvas. Secure. Structured. Ephemeral.

Vision

AI Is Moving Into Physical Space

AI began as chat. It evolved into copilots embedded in software. The next phase is ambient and spatial — systems that can borrow the nearest surface, present structured context, and step away the moment the job is done.

AI2X is the missing display infrastructure layer that makes this safe and repeatable: claim a screen, render deterministic content, then revoke with policy-bound guarantees.

Problem

AI has a voice. But it has no surface.

Trapped in chat windows

Output stays in a narrow UI, not in the workspace where decisions happen.

Single-device limits

Context does not move cleanly across wall displays, phones, panels, and wearables.

No spatial workflow

Teams lose clarity when AI cannot stage information across physical surfaces.

Why AI2X Exists

From trapped outputs to orchestrated surfaces

Without AI2X

AI trapped in chat UI
No spatial workflow across rooms
No deterministic rendering rules
No session governance

With AI2X

Dynamic display allocation
Structured templates for clarity
Policy-bound lifecycle
Ephemeral release model

Solution

AI2X as a Display Layer

Device discoverySecure session claimStructured renderingEphemeral release

Scenario

A Real Interaction Flow

1. Speak

User wearing AirPods asks the agent to review a document.

2. Detect

Agent recognizes that a visual review is required.

3. Claim

AI2X claims the nearest eligible display.

4. Render

Document appears via deterministic templates.

5. Confirm

User approves the decision and closes the task.

6. Revoke

AI2X revokes the session and clears the screen.

This flow maps directly to Claim -> Render -> Revoke and keeps the environment neutral between tasks.

Demo Film

Watch the display lifecycle in motion

Observe the orchestration sequence: a screen is claimed, content is rendered with structure, and the lease is revoked for ephemeral release.

Look for the clean handoff between Claim, Render, and Revoke.

Lifecycle

Claim -> Render -> Revoke

Claim

Agent requests lease and claims an eligible screen under policy checks.

Render

Deterministic templates render structured content with stable behavior.

Revoke

Session ends, content clears, display returns to neutral baseline.

Ambient AI -> Temporary Screen -> Review -> Revoke (Claim -> Render -> Revoke). Screen lights up -> content appears -> fades back to neutral.

Use Cases

Where AI2X is deployed

Contract Review

Ambient AI claims wall display, stages documents, then revokes.

Meeting Assist

Summaries and timelines appear across room displays and devices.

Ambient AI + Glasses

Wearable prompts pair with nearby displays for layered workflows.

Public Space AI Interaction

Policy-bound claims support kiosks and wayfinding in shared spaces.

Architecture

AI Agent -> OrchestrateDisplay -> MCP Service -> Display Gateway -> Runtime Display

AI2X enforces deterministic template contracts: no raw HTML injection, schema-driven rendering only, and risk-tier aware targeting for every display claim.

Delivery is AssignmentId-first, ensuring each render event is bound to a lease, a policy scope, and an auditable lifecycle.

Deterministic templates

Stable schema controls layout outcomes across display classes.

Risk-aware governance

Lease, renew, and revoke controls bound sessions for enterprise trust.

Engineering

Deterministic Display Contracts

Infrastructure-grade guarantees for display orchestration

AI2X enforces templateId-first rendering so every visual is validated against a contract before it appears.

No arbitrary UI injection. Structured rendering only.
Lease and renew lifecycle controls every session.
Session isolation per AssignmentId and policy scope.

Skill Installation

Install and test AI2X skill paths

Quick Start

# 1) Install openclaw
npm install -g openclaw
# 2) Add ai2x-skill
openclaw skill add ai2x-skill
# 3) Configure token + ctx path
openclaw config set AI2X_TOKEN="<YOUR_TOKEN>"
openclaw config set AI2X_CTX_PATH="/workspace/context"
# 4) Run test pushContent
openclaw run ai2x.pushContent --template contract.review --target wall-display-01

CLI

ai2x-cli auth login --token "<YOUR_TOKEN>"
ai2x-cli display discover --site "hq-floor-03"
ai2x-cli lease claim --display "wall-display-01" --ttl 300
ai2x-cli content push --display "wall-display-01" --template "contract.review" --data ./payload.json
ai2x-cli lease revoke --display "wall-display-01"

Docker

version: "3.9"
services:
  ai2x-mcp:
    image: your-org/ai2x-mcp:latest
    environment:
      - AI2X_TOKEN=<YOUR_TOKEN>
      - AI2X_CTX_PATH=/data/context
  ai2x-gateway:
    image: your-org/ai2x-gateway:latest
    depends_on: [ai2x-mcp]
# docker compose up -d

FAQ

lease: temporary claim
renew: extend an active claim before TTL
revoke: immediate release to neutral state
Rule: youtube.v2 must be full URL, e.g. https://www.youtube.com/watch?v=YOUR_VIDEO_ID

Team

Microsense Vision Co., Ltd.

Microsense Vision Co., Ltd. is an engineering-led AI infrastructure company delivering production-grade orchestration systems with deep hardware + software integration expertise for enterprise scale.

System architecture for multi-display orchestration
IoT device integration across heterogeneous hardware
Orchestration infrastructure and policy governance
Engineering and product integration
Mass production planning and rollout readiness

Engineering

Deterministic runtime contracts and session controls.

Product

Workflow-first interfaces for decision-grade output.

Infrastructure

Gateway operations, observability, and hardening.