AI-driven payroll engine for SMBs
Built for SMBs running Xero Payroll. Facial-recognition clock-in, fraud prevention, automated timesheet review, Xero sync.
What we built
AWS Rekognition facial recognition — Employee clock-in/out runs through AWS Rekognition. We enrol staff with multi-angle photos so the recognition step is reliable in poor light and across uniforms, and we keep the biometric template separate from the timecard record.
Multi-layer fraud prevention (GPS + WiFi + Bluetooth + biometric) — Every clock-in is verified by four independent signals: GPS location, the venue's WiFi network, a Bluetooth proximity beacon, and the facial-recognition match. A clock-in only counts when the layers agree, which makes buddy-punching and off-site punching essentially impossible.
Manager approval workflow with Xero sync — Managers review timesheets in a desktop dashboard with quality scoring and bulk approve/reject. Recognition metadata is preserved on the entry, so audit trails survive the approval step. Approved timesheets sync straight to Xero Payroll on a managed cadence.
See the full flow
One shift, end to end: face-scan clock-in, fraud checks, manager approval and Xero sync — in about ten seconds. Recorded on a synthetic demo org, Kauri Grove Care.
- Status
- Live in production
- First customer
- Norfolk Court Resthome
- Integration
- Xero Payroll synced
A closer look
How a shift gets paid.
- 01
Enrol
Multi-angle enrolment photos build the facial-recognition template.
- 02
Clock in
GPS + WiFi + Bluetooth + facial match all have to agree.
- 03
Review
Manager dashboard scores quality and approves in bulk.
- 04
Sync to Xero
Approved timesheets push straight to Xero Payroll.
Need an IT team that picks up?
Book a free 30-minute call. We'll tell you if we're a fit.