← All case studies
FACE CHECK-IN · Corporate
Attendance that takes a glance.
A tablet kiosk, in testing, that recognises staff by face and keeps attendance ready for payroll.
The work
What it actually looks like.
- Trigger Walk up to the tablet No card, no PIN.
- Step 1 The face is matched Compared with the enrolled staff, with a safety margin.
- Step 2 Attendance is logged Time and person, written once.
- Result The month is ready An export for payroll, no tallying.
In action
Four moments from one run.
Snapshots from the sample run above, one for each step. The data is made up; the steps are the real ones.
Before and after the automation.
| Task | Before | After |
|---|---|---|
| Check in | Sign a register | Look at the tablet |
| Month-end count | Tallied by hand | Exported |
| Wrong-person check-ins | Possible on paper | Designed to refuse when unsure; being tested |
How it is built
What it runs on.
Connects to
A staff database
Runs on
HubA tablet kiosk · A private server
Models
A face recognition model
Runs on
- A tablet kiosk
- A private server
Connects to
- A staff database
Models
- A face recognition model
The benefits
What it gives back.
No registers
Attendance is designed to be logged as staff walk in, with nothing to sign or tally.
Payroll-ready months
The month is exported for payroll instead of being counted by hand.
Unsure means no
It is built to ask again rather than guess, so a wrong check-in is not recorded.
Next up
WHATSAPP LEAD DESK · Field sales
Designed to answer, log and score every lead.
Read the case study →Want to discuss a project?