Platform architecture

iLive runs as a managed service you call from your apps and your backend. Your app — web or native — captures the user and talks to the iLive edge over HTTPS. Our backend does the verification work and returns a signed result. Your backend can also subscribe to webhooks for asynchronous outcomes.

High-level flow

        ┌───────────────────────────┐
        │   Your app (Web / iOS /   │
        │    Android / Flutter)     │
        │                           │
        │   • iLive SDK (optional)  │
        │   • Direct REST calls     │
        └──────────────┬────────────┘
                       │  HTTPS
                       ▼
        ┌───────────────────────────┐
        │        iLive edge         │
        │   REST API  •  SDK CDN    │
        └──────────────┬────────────┘
                       │
                       ▼
        ┌───────────────────────────┐
        │     iLive backend         │
        │                           │
        │   Liveness engine         │
        │   Face engine             │
        │   Collections & search    │
        └──────────────┬────────────┘
                       │
             webhooks  │  server-to-server API
                       ▼
        ┌───────────────────────────┐
        │     Your backend          │
        │   (records outcomes,      │
        │    drives next steps)     │
        └───────────────────────────┘

The SDK is optional but recommended for mobile and in-browser capture: it handles camera setup, session lifecycle, and result signing without you needing to build any of it. For server-side or back-office use cases, calling the REST API directly from your backend is fine.

Deployment options

SaaS (default). Hosted by iLive. Best for most teams — you get automatic updates, elastic capacity, and no infrastructure to run.

On-premise. For regulated environments that require all processing inside your perimeter, iLive can be deployed to your own infrastructure with the same REST API and SDKs. Contact us to discuss.

Data residency

On SaaS, you choose the region where your sessions and collections live; data stays in that region end-to-end. On-premise deployments keep everything within your network — nothing leaves your environment. Retention, deletion, and export are configurable per tenant.

Scale

Collections are built to hold millions of indexed faces with sub-second 1:N search. Liveness sessions complete in seconds and scale horizontally, so traffic spikes (onboarding campaigns, breach-response re-verifications) don't back up your queue. Rate limits are tenant-scoped and can be raised for production workloads — see Rate limits.