An employee time clock that runs on one Android phone — and never phones home.
FaceClock recognizes faces on-device, logs shifts offline, and exports CSV when you ask it to. No cloud account. No subscription. No biometric data leaves the phone.
Get it on Google Play
Why a face camera, and not a fingerprint or a PIN
Fingerprint readers wear out. PINs get shared. A camera mounted at the entrance does the boring thing well: it sees who walked in, marks the moment, and gets out of the way. The math runs locally on the device’s NPU, so the actual face data — embeddings, vectors, frames — never leaves the tablet sitting on your wall.
- Adds a person in under 30 seconds (3 reference photos, no setup wizard)
- Works without a network — recognized faces in roughly 200–400 ms
- Stores embeddings, not photos — you can review and delete them at any time
- Configurable confidence threshold for noisy lighting or shared shifts
- Falls back to a manual employee picker if recognition fails twice in a row
Shifts the way real schedules work
Time tracking software usually assumes a clean 9-to-5. Real schedules don’t. FaceClock handles split shifts, shifts that cross midnight, last-minute swaps, and the case where someone clocks out, comes back five minutes later, and the manager wants those merged into one continuous block.
- Shifts that cross midnight are stored as a single record, not two halves
- Auto-clock-out after a configurable timeout — useful when people forget to clock out at end of day
- Per-employee weekly hour caps with overtime flagging
- CSV export grouped by week, fortnight, or month — whichever your payroll runs on
- Edit a shift after the fact, with an audit log of what was changed and when
Built around the privacy laws you actually have to deal with
The biometric privacy landscape changed quickly: BIPA in Illinois, the EU’s GDPR Article 9, CCPA’s biometric clause, the new Texas CUBI updates. Most cloud time-clock vendors handle this by writing a 12-page privacy notice and asking employees to sign. We took a different approach — we made it impossible for the data to leave the phone in the first place.
- All face matching runs on-device. No biometric data is uploaded — there is no upload pipeline at all
- Employee record retention is configurable; default auto-cleans after 90 days
- Built-in consent capture flow with timestamp and signature, exportable as PDF
- Open source under Apache-2.0 — your auditor can inspect every line
Try it on a phone you already own.
FaceClock runs on most Android devices from the last five years. Install, register your team in a few minutes, and start logging shifts the same day. No account. No card.
Get it on Google Play