POST/v1/face/recognise
Face Recognition API
Detect and recognise faces — single or many — against your enrolled collections. Built on the same multi-face core running in production classrooms today.
early accessThe long-term goal of SirusAI is to expose its operational vision infrastructure through enterprise-grade APIs — so organisations and developers can build their own applications on intelligence that has already proved itself in the field. The rollout is progressive and deliberate: each API opens once the capability behind it has been validated in production.
The most trusted AI platforms followed the same arc: validate the model through real products first, then open the underlying capability to the world. SirusAI is following that arc deliberately.
Today, our recognition infrastructure earns its production record through the Classroom Intelligence platform — thousands of real faces, real lighting conditions, and real operational constraints, every single day. Every API we open will carry that production history behind it.
This is why our API layer rolls out progressively rather than all at once: we'd rather offer fewer endpoints that genuinely hold up than a long catalogue of capabilities that only work in demos.
The recognition core runs live inside the Classroom Intelligence platform, accumulating operational evidence across institutions.
Selected developers and organisations get first access to the recognition and liveness APIs, with direct engineering support.
Image intelligence, video intelligence, and CCTV intelligence APIs open in stages — each gated on proven reliability.
Operational automation APIs that let any organisation build vision-driven workflows on SirusAI infrastructure.
Six API families, opening progressively. Early-access partners help shape the interfaces before public release — if one of these is core to what you're building, talk to us now.
Detect and recognise faces — single or many — against your enrolled collections. Built on the same multi-face core running in production classrooms today.
early accessVerify that a face is a real, present person — not a photo, screen, or replay. The anti-fraud layer protecting every attendance record we process.
early accessBroader visual understanding of still images — scene context, environment awareness, and the building blocks of visual analytics.
in developmentUnderstanding across time: movement, events, and patterns in recorded or streaming video — the foundation for monitoring applications.
plannedLive intelligence on real CCTV feeds — the environment Show & Go was engineered for from day one. Recognition and awareness on the cameras already on your walls.
plannedThe destination: vision-triggered workflows. Define what your operation should do when the system sees something — and let the infrastructure act.
the goalThree commitments that hold for every endpoint we ever ship.
Every capability is validated through live production deployments before it reaches the API surface. You inherit the operational record, not just the model weights.
The infrastructure assumes low-quality streams, bad lighting, and crowded scenes by default — because that's what it was raised on. Clean input is a bonus, not a requirement.
Authentication, collections management, audit-friendly responses, and predictable behaviour at scale — designed for organisations that deploy, not just experiment.
Early-access partners shape the API surface and get direct engineering support from the team building the infrastructure. If vision capability is core to your product, this is the right time to start the conversation.