The Problem Is Supervision Density, Not Attendance

A security agency's structural problem is that guards work where supervisors are not. One field officer might cover fifteen or twenty sites. Attendance is recorded by the person whose attendance is being recorded, often at 10 p.m. at a gate nobody visits until morning.

Every workaround in this industry stems from that gap. A guard signs the register for a colleague who arrives an hour late. A relief guard covers two adjacent sites and both are marked present. The night shift is recorded as complete when the guard left at 4 a.m. None of these are exotic — they are the predictable consequence of a system where the only witness is the person being witnessed.

The commercial damage runs in both directions. You bill a client for a post that was unmanned, which becomes a contract problem when the client's own CCTV shows otherwise. Or you pay for hours that were not worked. Neither is fixed by a stricter register. Both are fixed by making the attendance record independently verifiable.

• One field officer covering fifteen-plus sites cannot witness attendance
• Register-based proxy marking is a structural outcome, not a discipline problem
• Billing a client for an unmanned post is a contract risk, not just a payroll leak
• The fix is an independently verifiable record, not tighter register rules

Why Hardware at Every Site Does Not Work Here

The instinctive answer is a biometric device at each post. For a security agency this rarely survives contact with reality. Client sites are not yours — installing hardware needs the client's permission, their power, and often their network. Contracts end and the device has to be recovered. New contracts start and you need a device before day one.

Guards are also frequently deployed to a different site than last week, and a device tied to a location does not follow the deployment. For agencies whose whole business is flexible deployment across changing client sites, a fixed installation is the wrong shape.

Face verification on the guard's own phone inverts the problem. It travels with the guard, works at a site you got yesterday, and needs no client permission or power. Attend Mitra runs face verification against an encrypted descriptor enrolled once, combined with a geo-fence defined per shift, so a guard deployed to a new site is covered as soon as the shift is created — not as soon as hardware is installed.

• Client sites need client permission, power, and network for any installation
• Devices must be recovered when contracts end and procured when they start
• Fixed hardware does not follow guards who move between deployments
• Phone-based face verification covers a new site the day you win the contract
• Geo-fence is defined per shift, so each post gets its own boundary

Night Shifts, Reliefs, and Rosters That Reflect Reality

Guard rosters are among the hardest scheduling problems in any industry. Twelve-hour shifts, shifts crossing midnight, rotating day and night patterns, weekly offs that move, and relief guards covering absences at short notice. A scheduling tool that assumes fixed weekday shifts is useless here.

Attend Mitra handles shifts that cross midnight without breaking the day's calculation, supports recurring patterns so a rotating cycle is defined once rather than re-entered weekly, and allows rosters to be saved as reusable presets. The weekly schedule stays in draft while you rearrange it and publishes to guards as a complete week, which matters when a roster changes four times before it is final.

The leave conflict warning is small and disproportionately useful. When you assign a guard to a shift they already have approved leave for, the system flags it during scheduling. In an industry where a missed assignment means an unmanned post and a client call, catching that on Thursday afternoon rather than Sunday night is the difference between an adjustment and an incident.

• Shifts crossing midnight calculated correctly for night posts
• Recurring rotation patterns defined once instead of assigned weekly
• Reusable roster presets for standard site staffing patterns
• Draft the week, rearrange freely, publish once as a complete roster
• Leave conflict warnings catch unmanned posts before the shift, not during it

Live Supervision Without Driving to Every Site

The manager app is where the supervision gap actually closes. A control room supervisor or field officer opens it and sees how many guards are clocked in against how many were scheduled, who is late, and who has not reported at all — filtered across every site at once rather than site by site.

The team map shows guards positioned as in-zone, out-of-zone, or tracking-only, so a post that should be manned and is not becomes visible immediately rather than at the next visit. A supervisor can request a live location refresh for a specific guard, which turns an unanswered phone call into a positive check. Live location is tied to being clocked in and stops at clock-out, and being explicit about that boundary with guards is what makes it acceptable rather than contentious.

Override control belongs in the same place. When a guard's phone fails or face matching struggles in poor light at a gate at 2 a.m., the supervisor issues a time-limited override code from the manager app instead of the guard bypassing verification. Every override is attributed to the supervisor who issued it, so exceptions stay visible. If overrides at one site rise week after week, that is a signal worth investigating rather than noise buried in a log.

• Scheduled versus clocked-in counts across all sites in one view
• Team map showing in-zone, out-of-zone, and tracking-only guards
• Request a live location refresh instead of calling an unanswered phone
• Time-limited override codes issued by supervisors, attributed and logged
• Rising override counts at one site are an early warning signal

Emergencies, Incidents, and the Payroll That Follows

Guard work carries genuine risk, and an attendance system that already knows who is on duty and where is the right place to handle it. Attend Mitra supports emergency alerts to all clocked-in staff or a selected group, with guards responding as safe or needing help, and their response locations visible to the supervisor. The same alert can be run as a drill, which is how you find out whether it works before you need it. Guards can also file incident reports from the app with status tracking, which builds a record for client conversations.

The payroll consequences are where verified attendance pays for itself. Guard payroll is dominated by overtime, night hours, and week-off working, and these are exactly the figures that get disputed. When hours come from verified clock-ins tied to identity and location, overtime is computed rather than negotiated, and loss of pay for unapproved absence is derived rather than argued.

The same record set supports client billing. When a client questions whether a post was manned on a specific night, you have a verified log with timestamps and locations rather than a signed register. That evidence is worth more than the attendance system costs, and it is the argument that usually decides the purchase.

• Emergency alerts to clocked-in guards with safe or need-help responses
• Run alerts as drills to test the process before a real incident
• Incident reports filed from the app with status tracking
• Overtime and night hours computed from verified logs, not negotiated
• Verified attendance evidence supports client billing disputes