Identity Agent
Multi-face recognition + watchlist matching on existing CCTV — door unlock, attendance, VIP routing, blacklist alert. No badges. No fingerprint queues.
“A swipe card proves someone had a card — not who walked in.”The face at the door, verified before it opens.
Everyone learns to live with the gap between “something happened” and “someone noticed.” We did not. Let the cameras do the watching and your people do the deciding — alerted the instant it matters, not hours later. From “that is just how it works” to “wait — it could work like this.”
Badges get shared. Fingerprints queue. Biometric kiosks break.
Most enterprises still run identity on badges, RFID cards, fingerprint kiosks and ID checks at the door. Badges get lent, lost, cloned. Fingerprint readers create morning queues. ID checks at the gate are theatrical — the guard glances, waves, and forgets. None of these systems answer 'who exactly was inside the facility at 02:47?' without a forensic exercise.
Worse, when a banned individual walks in — a known shoplifter into a store, a terminated employee back on the factory floor — your only chance of catching them is the human guard's memory of last Tuesday's briefing. That does not scale. That does not work on shift change. That does not work on the weekend night shift.
And the cost compounds. Most banks, malls, premium retailers and pharma facilities accept a small percentage of identity-related incidents each year as 'cost of doing business' because nobody can prove specifics. The fix is overdue.
Recognise. Verify. Decide. Unlock — or alert. Every event logged with evidence.
Identity Agent uses your existing CCTV to recognise faces against your VIP list, blacklist, and active employee roster. Per the VIZO361 datasheet, it covers two use cases — Multi-Face Recognition (Use Case 04) and Person of Interest identification (Use Case 05) — and requires 2 MP cameras minimum, with 1–3 days of zone deployment after watchlist enrolment.
On a match, the agent acts: door unlock via your existing access-control controller, attendance logged with timestamp and camera ID, blacklist alert pushed silently to security, VIP greeting routed to the manager. Every event is logged with face crop and camera evidence, searchable for years.
Privacy is built in. We do not store a biometric template — only a one-way vector hash that cannot be reverse-engineered. The vector you upload at enrolment never leaves your tenant. Deletion is real deletion. Our default configuration is DPDP-aligned (India), GDPR-compatible (EU), Oman PDPL-aligned, with data-residency selection.
Known limitations (from the VIZO361 datasheet): performance drops with low light, face masks, or angled faces. POI matching requires high-quality enrolled face data. For high-stakes identity-only decisions, the agent flags potential matches for human verification rather than acting unilaterally.
How Identity Agent works — end-to-end
From the moment a frame leaves your camera to the moment your hardware reacts. No black box — every step explained.
Face capture
Existing CCTV at 2 MP+ (datasheet minimum). Multi-scale detector handles small faces, partial occlusion, side angles.
Quality + liveness
Image quality score gates further processing. Liveness model defeats common photo and replay attacks.
Vector + match
Face encoded to a one-way vector, matched against your enrolled tenant gallery (employees / VIPs / blacklist) using cosine similarity.
Policy decision
Match + role + time-window + camera context → policy decides: unlock / log only / silent alert / PA-page / manager-notify.
Hardware actuation
Door controller relay or REST call to access-control. Parallel events to attendance system, watchlist alert, VIP greeting.
Detection without action is just an alarm. We close the loop.
Every Identity Agent event can trigger your existing on-site hardware. Wired or wireless. PLC, contactor, BACnet, MQTT or webhook — your choice.
Door / Turnstile Unlock
Integration
Dry-contact relay or REST to existing access-control (HID, Lenel, Honeywell, Salto, Suprema, ZKTeco, custom). Wiegand / OSDP bridge supported.
Outcome
Authorised face → door opens. Hands-free. No badge, no fingerprint queue.
Attendance Auto-Log
Integration
REST / webhook to HRMS (Zoho People, Keka, GreytHR, BambooHR, SAP SuccessFactors, custom).
Outcome
Clock-in with timestamp, camera, and face crop. Eliminates buddy-punching. Audit-grade attendance.
Blacklist Silent Alert
Integration
WhatsApp, SMS, Slack, Teams, custom webhook to security operations.
Outcome
Banned individual enters → silent alert with face crop, camera and timestamp. No alarm to alert the subject.
VIP Welcome Routing
Integration
Floor staff WhatsApp, manager Slack ping, CRM lookup (Salesforce, Zoho, Dynamics).
Outcome
Premium customer recognised at door → preferred manager notified with relationship summary before they reach the sales floor.
Visitor Pre-Approval
Integration
Visitor-management system (Envoy, SwipedOn, custom). API webhook with photo on host invite.
Outcome
Pre-registered visitor walks in → face matched against invite → host auto-notified + temporary access granted for booked duration only.
Time + Attendance + Payroll
Integration
Direct push to payroll (Zoho Payroll, SAP, Ramco, custom). Reconciled with shift roster.
Outcome
End-of-month payroll-ready data. Statutory compliance trail for PF / ESI / WPS.
Where Identity Agent is deployed
Specific industries, specific scenarios, specific outcomes — not generic marketing claims.
BFSI Branches
VIP recognition + watchlist + teller-area attribution
Premium customers recognised at door, branch manager notified. Watchlist match against known fraudsters or impersonators → silent alert. Vault-access logged with face + timestamp.
Corporate Campuses
Visitor mgmt, contractor access, VIP visits
Pre-registered visitors auto-enter via the same lobby cameras. Contractor permits enforce per-zone and per-time access. Executive visits auto-route to host.
Healthcare
Restricted-zone access — NICU, OR, pharmacy
Only authorised clinical staff enter ICU / NICU / OR / pharmacy. Patient-floor visitor management. Anti-abduction policy for paediatric zones.
Premium Retail & Auto Showrooms
VIP service, loss prevention
High-value customer recognised → senior advisor routed in seconds. Known shoplifter walks in → silent alert with visible store-team presence.
Critical Infrastructure & Defence
Restricted-zone watchlist enforcement
Pre-cleared personnel verified at zone entries. Non-cleared individual detected → silent alert + door lock + security dispatched. Air-gapped on-prem deployment supported.
Manufacturing & Pharma
Clean-room access, contractor compliance
Clean-room entry tied to qualification status pulled live from training system. Contractor permits enforced per-zone and per-time.
Technical specifications
The numbers we publish in proposals. Detailed accuracy targets, latency budgets, and integration scope.
| Datasheet reference | Use Cases 04 + 05 — Multi-Face Recognition + Person of Interest (POI) Identification |
| Deployment timeline | Approximately 1–3 days per zone, per the datasheet (plus watchlist enrolment time) |
| Minimum camera resolution | 2 MP |
| Training required | Minimal once enrolled; initial face-database upload required |
| Liveness detection | Designed to defeat common photo and replay attacks; benchmarked per deployment |
| Mask / occlusion tolerance | Eye + forehead recognition supported with documented accuracy degradation; benchmarked per site |
| Compatible cameras | Standard RTSP / ONVIF IP cameras, 2 MP–8 MP |
| Scale | Up to 500+ concurrent camera streams in the enterprise tier |
| Privacy & storage | One-way vector hash stored. No reversible biometric template. Vector hashes do not leave tenant. DPDP / GDPR / Oman PDPL aligned. |
| Deployment | On-prem (air-gap supported), cloud, hybrid. Multi-site federation with central enrolment and edge match. |
| Benchmark accuracy | Site-specific — measured and documented during the 30-day pilot |
Eliminate buddy-punching. Accelerate VIP service. Catch repeat offenders before they reach the floor.
Direct labour: replacing badge and fingerprint readers with face access typically allows redeployment of reception / security personnel from the queue to active duty. The specific count and value depends on your sites, shifts, and current arrangement — benchmarked during the pilot.
Loss prevention: BFSI customers use the watchlist for branch-fraud reduction. Premium retailers protect revenue from previously-tolerated repeat shoplifters. Hospitals reduce restricted-zone breach incidents. The specific size of the gain depends entirely on your current incident rate and is benchmarked during the pilot.
Top-line: VIP recognition at the door enables senior advisors to greet premium customers before they walk past the queue. The lift on premium conversion at your specific brand is measured during the pilot.
A real-world example
Customer name withheld per NDA. Full reference available on request.
Reference customer (named under NDA)
Proeffico has deployed Identity Agent across reference customer sites in BFSI, healthcare, premium retail and corporate campuses. Verified outcomes are available under NDA on request. Site-specific accuracy and incident-reduction are benchmarked during your 30-day pilot on your actual cameras and shared in a signed pilot report before any commercial commitment.
Pilot
Real numbers benchmarked on your site, in your signed pilot report
Identity Agent vs the traditional way
Most facilities still run on the old playbook. Here's what changes.
The traditional way
- ✗Badge / RFID — gets lent, lost, cloned. No verification badge belongs to bearer.
- ✗Fingerprint kiosk — slow per-person processing. Hygiene concerns.
- ✗Watchlist enforcement relies on guard memory of weekly briefing
- ✗Attendance via punch-in apps → buddy-punching, ghost employees
- ✗VIP arrives → recognised by chance by whichever staff happen to look up
With Identity Agent
- Biometric verification against face. Cannot be lent. Liveness-checked.
- Contactless face recognition. No queues. No touch.
- Watchlist enforced 24/7 by every camera. Silent alert on any match.
- Face-verified attendance. Buddy-punching impossible. Ghost employees surfaced in audit.
- VIP recognised at door → preferred manager notified with full context before customer reaches them.
Questions buyers ask before they sign
Is this legal in India / GCC / EU?
+
Yes, when configured correctly. Identity Agent stores only a one-way vector hash, and your tenant retains control of consent, enrolment and retention. Our default configuration is DPDP-aligned (India), GDPR-compatible (EU), and Oman PDPL-aligned, with data-residency selection. We provide the DPIA template, data-flow document, and consent forms during deployment.
How does liveness detection work?
+
Multi-cue: eye micro-movement, head-pose change over time, depth cues from monocular geometry, and skin-texture analysis. The specific accuracy against printed-photo / screen-replay / 3D-mask attacks is benchmarked during the pilot. For high-security deployments we add a second camera angle and an active prompt.
Will it work through masks and hats?
+
Per the datasheet, performance drops with face masks and angled faces. Eye + forehead recognition is supported with documented accuracy degradation, which is benchmarked at your site during the pilot. For exceptionally challenging conditions, badge fallback is recommended.
What if my employee photos are 5 years old?
+
We re-enrol on first sighting — when an older photo matches a new face capture with high confidence, the system upgrades the enrolment automatically (with optional admin review). This keeps the gallery current without forcing a re-enrolment day.
Can it integrate with our existing access-control system?
+
Yes — we support HID, Lenel, Honeywell, Salto, Suprema, ZKTeco, ESSL, Matrix, Spintly and dozens more via Wiegand bridge, OSDP, REST API or dry-contact relay. Confirmed during the site survey.
What happens if the camera misidentifies someone?
+
Every match is logged for review. The system supports an admin-override review flow for any disputed event, and the agent learns from corrections. False-positive thresholds are tuned during the pilot.
Can I run it air-gapped?
+
Yes. For defence, regulated finance and clinical-trial sites we deploy fully air-gapped on-prem. No internet required for enrolment, match, decision or actuation. Software updates via controlled SCP / USB.
What does it cost?
+
VIZO361 is sold on a Private Offer model — pricing reflects enrolled head count, cameras, integrations and deployment mode. A 30-day pilot on 2 cameras and a small enrolment is included.
From the live VIZO361° platform — not mockups
Screenshots and alert clips captured from the deployed product. Customer identifiers have been blurred where present.
Who built this — and who reviewed it
Last reviewed: 2026-05-22Product Lead
Utkarsh Yadav
AI / ML Lead, VIZO361° (Facial Recognition + Liveness Detection)
Leads the facial-recognition, liveness-detection, and identity workflows in VIZO361. Designs the agent architectures behind Identity, Footfall Analyst, and PPE Compliance.
Product Lead
Swastik Vaish
AI Engineer, VIZO361° (Fire & Smoke, Cash Detection)
Leads the safety + loss-prevention model family in VIZO361 — including Fire Watch, Cash Counter, Guard Watcher, and Crowd Density. Owns model accuracy benchmarking during pilots.
Reviewed By
Saurabh Agarwal
Founder & CEO, Proeffico Solutions Pvt Ltd
Founded Proeffico in 2018. Reviews every product narrative for accuracy against the VIZO361 datasheet, customer reference behaviour, and Proeffico's published trust commitments.
Profile →All capability claims on this page are anchored to the VIZO361 Product Datasheet (2025). Site-specific accuracy, latency, and ROI figures are benchmarked during a 30-day pilot on your cameras and shared in a signed pilot report — not published as universal numbers.
Explore other Vision Agents
Private Offer · No published list price
See Identity Agent on your cameras. 30-day pilot. NDA. Real conditions.
Two-camera pilot on your site. Real models. Real lighting. Real workflows. Quote within 5 working days of discovery.