Technology
Powered by Casual Mode™
Powered by Casual Mode — our proprietary architecture that delivers sub-second AI responses with 95% token reduction through local pre-training and edge computing.

Core Architecture
Casual Mode™ Technology
Casual Mode eliminates cloud dependency through local pre-training, delivering sub-second responses with 95% token reduction and zero idle consumption. AI processing runs entirely on-device, keeping conversations fast, private, and cost-efficient.

Intelligence-Display Separation
Decouples AI processing from visual rendering, allowing each component to scale independently for optimal performance and resource utilization.
Zero Idle Consumption
The LLM sleeps when no user is detected, resulting in zero token cost during idle periods — dramatically reducing operational expenses.
Local Rendering
On-device NPU handles all visual rendering locally, eliminating the need for cloud streaming and ensuring instant avatar responsiveness.
Pre-Trained Actions
Gestures, expressions, and behaviors are pre-trained and execute locally, enabling natural avatar interactions without network latency.
Trust & Compliance
Enterprise-Grade Security
Privacy-First
No cloud video streaming or storage. All visual processing stays on-device, ensuring complete privacy for every interaction.
Data Sovereignty
Deploy on-premises, in the cloud, or in a hybrid configuration — you choose where your data lives and how it flows.
Compliance
Built to meet PDPA, GDPR, ISO 27001, and SOC 2 standards for regulated industries and government deployments.
Encryption
TLS 1.3 for data in transit, end-to-end encryption for AI conversations, and AES-256 at-rest encryption for stored data.
Connectivity
Enterprise Integration
Connect VoxHolo to your existing systems with robust API connectivity and enterprise-grade authentication.
- RESTful & GraphQL APIs
- OAuth 2.0 & Enterprise SSO
- ERP, CRM & IoT Integration
- Webhook & Real-Time Events
// Initialize VoxHolo client
const voxholo = new VoxHolo({
apiKey: process.env.VOXHOLO_KEY,
region: "ap-southeast-1",
});
// Deploy an avatar with custom brain
const avatar = await voxholo.avatars.create({
name: "Concierge",
language: ["en", "zh", "ms"],
brain: "hospitality-v2",
mode: "casual",
});
// Stream real-time events
avatar.on("conversation", (event) => {
console.log(event.transcript);
});See it in Action
Experience Casual Mode technology first-hand. Book a live demo with our team and see how VoxHolo transforms customer engagement.