Sovereign. Edge-Native. Face Recognition.
On mobile, tablet and CCTV. Works offline.
One NIST-benchmarked engine, running where your data lives: on-device for Android and iOS phones and tablets, and on-premise for the CCTV cameras you already have. No cloud dependency.
On-device: mobile and tablet
SDK for Android and iOS. 1:1 verification, 1:N identification and passive liveness, fully offline.
On-premise: CCTV and video
NBOX edge appliance for existing RTSP/ONVIF cameras. Live alerts and video forensics, on-site.
Deployed in India, UAE, Saudi Arabia, Kuwait, Bahrain, Malaysia, Australia, Kenya and Nigeria.
Face recognition for the work teams actually do
Works offline. Runs on the device.
Face recognition attendance
Staff clock-in and clock-out for multi-site teams in the Gulf, India, Malaysia, Australia and Africa.
Learn moreStudent attendance
Mark a whole class in one scan, with no roll call.
Learn moreExam candidate verification
Check every candidate at the door, even in centres with weak internet, from India to Kenya and Nigeria.
Learn moreVoter verification
Used in the Bihar panchayat elections to block impersonation.
Learn moreFace verification for HRMS and payroll software
Embed the SDK to stop buddy punching in time-clock products.
Learn moreOffline face recognition on mobile and tablet
Matching on the device when connectivity is poor.
Learn moreCCTV face recognition on existing cameras
Add face recognition to the RTSP/ONVIF cameras you already have, on-premise.
Learn moreLive watchlist alerts on CCTV
Real-time alerts for watchlisted faces at entry points, perimeters and public sites.
Learn moreVideo forensics and face search
Search hours of recorded CCTV footage for a person in minutes.
Learn moreBuilt for
HRMS, exam platforms & identity verification providers
Embed face recognition via SDK or API. No AI team required. We're your biometric engine.
See exam verificationSystem integrators & CCTV operators
Add Vision AI to any existing camera setup without rip-and-replace. Resell to your clients. White-label ready.
Security consultants & solution builders
Specify FaceTagr in your design. We handle the AI. You deliver the complete solution.
Edge AI on phones, tablets and devices
Face recognition SDK with passive liveness. On the device, not in the cloud.
Embed 1:1 verification, 1:N identification and passive liveness in Android, iOS, Flutter or React Native apps. Recognition and liveness run locally and work offline, using FaceTagr's own NIST-benchmarked engine.
Passive liveness
No blink, head turn or challenge prompts. Evaluate photo and screen replay protection from a normal camera.
Works offline
Recognition and liveness do not need a connection. Your app can sync agreed records later.
Standard phones and tablets
Native Android and iOS, Flutter and React Native. Mobile testing spans 18,000+ Android device models; validate your own devices.
Choose fully offline matching or a server-connected design. Define enrolment, storage and sync for your application.
Edge face recognition SDK and FaceTyck
For HRMS providers, payroll companies, exam platforms, and any product team that needs face recognition as a feature - not a project.
Face Recognition SDK & API
Edge-native SDK for Android and iOS with on-device 1:1 verification, 1:N identification, and passive liveness. Fully offline biometric workflows, Flutter and React Native support, plus a RESTful API. Volume licensing available.
FaceTyck Attendance
FaceTagr's face recognition engine is tested by NIST FRTE, with up to 99.91% match accuracy.
FaceTyck is FaceTagr's app for all on-device face verification and recognition: staff and student attendance, exam checks, works offline.
Out-of-box HRMS Integrations
The Runtime Behind Every Deployment
NBOX is not just a hardware box. FaceTagr is a Vision AI runtime - the same NIST top-1% accuracy engine running across fixed CCTV infrastructure, Android field devices, drone footage, and Meta AI Smart Glasses. Camera-agnostic. Hardware-agnostic. Offline. The same engine powers every NBOX unit and every SDK integration.
Any IP camera
RTSP / ONVIF. Any brand, any age, any resolution.
Mobile and tablet devices
FaceTyck Attendance and SDK: Android and iOS. FaceTyck Attendance needs 2GB RAM minimum; the SDK needs 1GB RAM minimum. Android testing spans 18,000+ models.
Drone & aerial
Face recognition, people detection, object detection on drone footage.
Meta AI Glasses
Wearable hands-free field ID. Android phone as edge compute. No cloud.
Camera-agnostic. Hardware-agnostic. Up to 99.91% match accuracy in NIST FRTE testing - top 1% band in NIST FRTE 1:N identification. ISO 27001 certified. Hosted on SOC 2 Type II-certified infrastructure. Deployed across 9 countries.
How Face Recognition Works
Detect, align, encode, match - the full biometric pipeline in under 70 milliseconds. Every step runs on the edge, so raw biometric data never leaves the premise.
- →Face detection on any RTSP / ONVIF stream or mobile camera
- →Passive liveness check - no blink, no turn, no prompts
- →On-device encoding into a non-reversible face template
- →1:1 verification or 1:N identification, fully offline
NBOX for CCTV
NBOX adds face recognition to the CCTV cameras you already have.
Two variants based on your deployment need.
NBOX Real-Time
Connects to your existing CCTV via RTSP/ONVIF. Processes every frame. Alerts, watchlists, attendance - live. Self-configurable dashboards for any industry. RESTful API to push events to any downstream system.
Works fully offline and on-premise. Hardware-agnostic. Any brand camera. Install in minutes.
Starting from ₹5 Lakhs / ~$6,000 USD - including hardware - ships worldwide
NBOX Forensics
One device. No live cameras needed. Upload video files or images from any source - multiple feeds at once. Run face recognition, object detection, and natural language queries across all of it.
"Find all instances where someone in a blue shirt was near the entrance between 6pm and 9pm across all camera feeds."
Ideal for police stations, campuses, corporate security - when something happens and you need answers, not real-time.
Learn moreWhy Integrators Choose FaceTagr
Camera-agnostic. Hardware-agnostic. No lock-in.
NBOX connects to any IP camera brand via RTSP or ONVIF. No proprietary hardware. No vendor tie-in. Your clients keep their existing cameras.
Works offline in extreme environments
Deployed in operating mines, highway construction corridors, and air-gapped government networks. Full functionality without internet. This is where other systems fail.
Matches 30-year-old photographs accurately
Low-resolution voter ID photos from decades ago matched reliably against present-day faces. Proven across Bihar state panchayat elections - 4,000+ booths, approximately 10,000 fraudulent attempts blocked.
Complete ecosystem - not just a plugin
Configuration console, dashboards, reports, mobile alert apps for Android and iOS, RESTful API. Your client gets a finished working system. DIY deployment - no professional services required.
Five form factors. One platform.
CCTV edge unit, Android mobile, drone, Meta AI Smart Glasses, cloud API. Switch deployment contexts without re-integrating. One engine, any environment.
Deployed and Proven
Bihar Panchayat Elections
- Problem
- Bogus voting across rural booths, with only decades-old low-quality voter ID photographs to match against.
- Deployment
- 4,000+ booths covered in a single-day statewide rollout on mobile devices, fully offline.
- Result
- 5M+ voters verified against decades-old ID photographs at national scale.
Multi-Site Enterprise Attendance
- Problem
- Proxy attendance and disputed clock-ins across manufacturing, mining, and highway construction sites with unreliable connectivity.
- Deployment
- Mounted and handheld devices plus existing CCTV, one enrollment shared across channels, offline-first with HRMS sync.
- Result
- Attendance disputes reduced to near zero and payroll cut-offs met without manual reconciliation.
HRMS and Payroll Platform Distribution
- Problem
- HR and payroll platforms needed biometric attendance from a vendor that understood payroll and workforce realities, not just an image API.
- Deployment
- FaceTagr attendance embedded and shipped inside partner HRMS, WFM, and payroll platforms through SDK and API, on Android and iOS.
- Result
- Live production attendance inside partner platforms across aviation services, hospitality, infrastructure, and retail accounts.
Deployed at
A selection of enterprise and government deployments across aviation, manufacturing, healthcare, energy, retail, and public administration - not an exhaustive list.
Different faces. Different conditions. Test on your own users.
FaceTagr supports deployments across countries, cameras and device types. NIST benchmarking provides independent evidence, not a guarantee for every population. A pilot should cover your actual lighting, image quality, users and exception process.
Across locations
Choose device, on-premise or regional hosting and document where data is held.
Across conditions
Measure false matches and missed matches in bright sun, dim rooms and your actual enrolment images.
Across devices
Pilot your supported phones, tablets or existing CCTV instead of assuming identical performance.
Frequently Asked Questions
FaceTagr is not a camera product. It is a Vision AI runtime - the same NIST top-1% engine running across a CCTV edge unit, an Android phone, a drone, and Meta AI Smart Glasses in the field. Every other system in this space is cloud-dependent or hardware-locked. FaceTagr is neither. Five form factors. Nine countries. One billion verifications in production.
Investor and partnership enquiries → contact@facetagr.com