Watchlists
Separate lists for staff, contractors, visitors and persons of interest, each with its own rules.
Face Recognition System
Recognition that stays inside the rules
Watchlists, access control and visitor logs — with consent, retention and audit built into the workflow rather than bolted on afterwards.
Home › Face Recognition System
Overview
A capability that has to be governed, not just deployed
Face recognition is the easiest module to sell and the easiest to get wrong. Enrol without consent, keep templates forever, give everyone access to the log, and a security improvement turns into a liability.
We deploy it with the controls first: who may enrol, who may search, how long a template lives, and a log that shows every match and every lookup. If your use case does not need identity, we will tell you to use anonymous counting instead.
- Consent and enrolment workflow, not silent capture
- Retention rules per watchlist, enforced automatically
- Every search and match written to an audit log
- Templates stored on infrastructure you control
Capabilities
What it does
Every item below is in the product today, not on a roadmap slide.
Access control
Face as a second factor at doors and gates, integrated with your existing controller.
Visitor management
Pre-registration, badge issue and automatic expiry when the visit window closes.
Time & attendance
Contactless attendance for sites where cards get shared and registers get signed in bulk.
Audit & governance
Who searched, who matched, who exported — every action recorded and reviewable.
Retention control
Template and image lifetimes set per list, with automatic deletion when the period ends.
How we deploy
From survey to steady state
Four stages, and we will tell you at the end of any of them if the rest is not worth doing.
Assess
We check whether identity is genuinely required, or whether anonymous analytics solves the same problem.
Govern
Consent flow, retention periods, roles and audit requirements agreed and written down before enrolment starts.
Enrol
Staff, contractors and visitors enrolled through a controlled workflow with quality checks.
Operate
Matching in production, with periodic reviews of the audit log and list hygiene.
Specifications
The practical detail
- Matching1:N watchlist search and 1:1 verification at access points
- StorageTemplates and images stay on your infrastructure, with a retention period set per watchlist.
- Access pointsIntegrates with existing access controllers and turnstiles over Wiegand, relay or API.
- AuditEvery enrol, search, match and export logged with user, time and reason
- ComplianceDesigned around India's DPDP Act obligations, including consent, retention and erasure.
- FallbackCard or PIN fallback so a failed match never traps someone at a door
Questions
Frequently asked
Is this legal to deploy in India?
Face data is sensitive personal data under the DPDP Act, which means lawful purpose, notice, consent for most uses, and defined retention. The system is built to support those obligations, but the legal responsibility sits with you as the operator — have your counsel review the specific deployment. We are engineers, not your lawyers.
What accuracy does it achieve?
It depends on camera angle, lighting, list size and how tight you set the threshold. A tighter threshold means fewer false matches and more missed ones — that trade-off is a policy decision we will walk you through, not a number we quote blind.
Can we use it without storing images?
Yes. Templates can be stored without retaining the source image, which is usually the better default.
What if we only need to count people?
Then use anonymous analytics instead. Fewer obligations, less risk, same operational answer.
Read more
More on face recognition
Face recognition attendance systemVideo analytics use casesCampus CCTV
Want to see it on your own cameras?
Send us a camera list and what you are trying to catch. We will tell you what is realistic before anyone talks price.