Glow Homes deploys around 80 staff across six zones at two Mahindra locations in Bengaluru. The site supervisor is appointed by the client and is on the floor, not at a desk. The field app is built around them.
Mobile-only. No laptop required. Works on Android and iOS phones supervisors already own.
Face-scan attendance. Punch in and punch out from the supervisor's phone. The capture path is wired end to end; frame capture, liveness and face matching are modelled and remain a contract item.
Overtime at source. Raised at site, approved by head office, flows into the same month's payroll.
Service requests. Extra pantry, replacement housekeeping, deep clean. Tracked to closure.
Big icons, big buttons. Designed for use with gloved hands, in low-light corridors.
Vernacular. Hindi and Kannada are an add-on, not part of Essential V1.
9:42●●● ▮▮▮ 89%
GOOD MORNING
Loading...
Loading shift...
11
Present
1
Absent
1
Late
12
Roll
Face-scan attendance×
Camera not opened. Capture is modelled in this demo.
No frame is captured and no face template is matched. The punch is recorded against the
selected employee so the path from site to the attendance grid is real end to end.
Real capture with liveness is a contract item.
Quick actions
◉
Face-scan Attendance
⚠
Report Issue
Today's tasks · 2/7 done
Action recorded.
Try it
This is a live preview.
Every quick-action button on the phone calls a real API endpoint on the mock server. The toast confirmation shows the server response.
Tap "Face-scan Attendance" — pick an employee and punch in or out. The scan is staged and liveness is modelled, but the punch is recorded and shows up on the attendance day grid in the ops console.
Tap a task — toggle status; progress updates.
In production, the app would have offline-first sync (supervisors aren't always on WiFi), push notifications for owner escalations, and a one-tap shift-handover workflow.