Sawaed
HR and payroll platform for Saudi companies, from field check-in to the Mudad file
-
At a glance
Sector
Human resources
Platforms
Web · iOS · Android · WhatsApp
Modules
9
Integrations
8
Screens
4
Total duration
23 weeks
Sawaed records office and field attendance from the phone, inside the site geofence, with a selfie and liveness check. It runs shifts, leave and approval chains, calculates payroll and GOSI contributions, and prepares the wage-protection file for Mudad, while tracking contracts in Qiwa, iqamas in Muqeem and Saudisation in Nitaqat. A WhatsApp assistant answers staff questions and files their leave requests.

-
Challenge
-
01
Field attendance lives on paper sheets and WhatsApp photos, with no proof of who actually reached the site.
-
02
Payroll is built in spreadsheets, and wages that differ from GOSI records surface only after the Mudad upload, when compliance has already dropped.
-
03
Iqamas and exit re-entry visas lapse unnoticed, and no one can tell how a hire or a resignation will move the company’s Nitaqat band.
-
04
Leave requests wait in a manager’s inbox, and repeated questions about balances and policy eat the HR team’s day.
Who it is built for
-
Employee
Checks in, requests leave and sees balances from the phone or WhatsApp
-
Field supervisor
Sees who is on each site and approves requests from the field
-
HR specialist
Files, contracts and documents kept current from Qiwa and Muqeem
-
Payroll accountant
A correct run from attendance to the Mudad file and the journal entry
-
HR director
Sight of Saudisation, cost and compliance before they turn into violations
-
Modules
What the system does
Connected modules working on one set of data.
-
01
Geofenced attendance
Mobile check-in inside the site geofence with a selfie liveness check
-
02
Shifts & rosters
Fixed and rotating shifts, Ramadan hours and overtime under the Labour Law
-
03
Leave & permissions
Balances and requests with approval chains by department, type and duration
-
04
Payroll & GOSI
Monthly run with allowances, deductions, GOSI contributions and a journal entry
-
05
Wage protection
Builds the Mudad payroll file and tracks its status and compliance rate
-
06
Contracts & documents
Contract sync with Qiwa, plus iqama and visa alerts from Muqeem
-
07
End of service
Benefit calculated under the Labour Law by reason of exit, then the final settlement
-
08
Saudisation board
Current ratio and band, with what-if for hires and exits on Nitaqat
-
09
Self-service & assistant
An employee app and a WhatsApp assistant that answers questions and files requests
Where the AI works
-
AI · 01
A face-match and liveness model compares the check-in selfie with the enrolled photo and rejects pictures of screens or printouts.
-
AI · 02
A private in-Kingdom language model answers from company policy and the Labour Law, citing the article, and turns a message like “I need Sunday to Thursday off” into a request ready for approval.
-
AI · 03
A pre-approval check compares each salary with last month and with the wage registered at GOSI, flagging unusual variances before the Mudad file is built.
-
AI · 04
A Saudisation forecast projects how planned hires and expected exits move the 26-week weighted average used by Nitaqat.
-
Screens
The system as the client's team and customers use it.



-
Architecture
Independent services, in-Kingdom data, and integrations over documented APIs.
Channels
- Employee app: iOS & Android
- WhatsApp assistant: Staff questions & requests
- HR & payroll console: Web app for HR teams
Edge
- API gateway: Routing & rate limits
- Identity: Nafath, roles & trusted devices
Services
- Attendance service: Check-ins & geofences
- People & leave: Records, leave & requests
- Workflow engine: Approval chains & scheduled jobs
- Payroll engine: Payroll, GOSI & end of service
- Compliance service: Documents, Nitaqat & deadlines
AI
- Face & liveness: Face match & liveness
- HR assistant LLM: Private in-Kingdom model
- Payroll checks: Variance detection before approval
Data
- PostgreSQL · PostGIS: Records & geofences
- Policy index: Vector index of policy & Labour Law
- Object storage: Photos & documents in-Kingdom
Integrations
- Mudad: Wage protection (WPS)
- GOSI: Social insurance
- Qiwa: Contracts & Nitaqat
- Muqeem: Iqamas & visas
- Bank payroll: Salary transfers
Connections
- Employee app to API gateway
- WhatsApp assistant to API gateway
- HR & payroll console to API gateway
- API gateway to Identity
- API gateway to Attendance service: REST
- API gateway to People & leave
- API gateway to Payroll engine
- API gateway to Compliance service
- Attendance service to Payroll engine: events
- People & leave to Workflow engine
- Attendance service to Face & liveness
- People & leave to HR assistant LLM
- Payroll engine to Payroll checks
- HR assistant LLM to Policy index
- Attendance service to PostgreSQL · PostGIS
- Attendance service to Object storage
- Payroll engine to PostgreSQL · PostGIS
- Compliance service to PostgreSQL · PostGIS
- Payroll engine to Mudad: WPS
- Payroll engine to GOSI
- Payroll engine to Bank payroll
- Compliance service to Qiwa
- Compliance service to Muqeem
Stack
- Next.js
- NestJS
- Flutter
- PostgreSQL · PostGIS
- Redis
- Temporal
- Python · FastAPI
- Kubernetes
-
Workflow
From check-in to payslip
-
01
Geofenced check-in
EmployeeEmployee app · Face & liveness
-
02
Exceptions & late arrivals
SupervisorAttendance service
-
03
Approve leave & overtime
Manager, then HRWorkflow engine · WhatsApp
-
04
Close attendance & calculate payroll
AutomaticPayroll engine
-
05
Variance check & GOSI match
AI, then payroll accountantPayroll checks · GOSI
-
06
Approve the run
Finance managerHR & payroll console
-
07
Bank transfer & Mudad upload
AutomaticBank payroll · Mudad
-
08
Payslips & journal entry
AutomaticEmployee app · ERP
-
Data & API
The core entities and relations, and a live example from the API.
Employee
- iduuidPK
- emp_notextunique
- national_idtext
- nationalitychar(2)
- is_saudiboolean
- site_iduuidFK
- manager_iduuidFK
- hired_atdate
- statusenum
Contract
- iduuidPK
- employee_iduuidFK
- typeenum(fixed,
open) - basicnumeric
- allowancesjsonb
- qiwa_reftext
- starts_ondate
- ends_ondate
- eosb_accruednumeric
Site
- iduuidPK
- nametext
- citytext
- centergeography(point)
- radius_msmallint
- requires_selfieboolean
CheckIn
- idbigserialPK
- employee_iduuidFK
- site_iduuidFK
- typeenum(in,
out) - attimestamptz
- pointgeography
- distance_mreal
- livenessreal
- statusenum
LeaveRequest
- iduuidPK
- codetextunique
- employee_iduuidFK
- typeenum
- from_datedate
- to_datedate
- daysnumeric
- statusenum
- channelenum(app,
whatsapp, web)
PayrollRun
- iduuidPK
- periodtext
- statusenum
- grossnumeric
- gosi_employernumeric
- netnumeric
- mudad_file_reftext
- approved_byuuidFK
Payslip
- iduuidPK
- run_iduuidFK
- employee_iduuidFK
- earningsjsonb
- deductionsjsonb
- gosi_employeenumeric
- netnumeric
- ibantext
Document
- iduuidPK
- employee_iduuidFK
- kindenum(iqama,
visa, passport, permit) - numbertext
- expires_ondate
- sourceenum(muqeem,
qiwa, upload) - alert_stateenum
- Employee 1—∞ Contract
- Site 1—∞ Employee
- Site 1—∞ CheckIn
- Employee 1—∞ CheckIn
- Employee 1—∞ LeaveRequest
- Employee 1—∞ Document
- PayrollRun 1—∞ Payslip
- Employee 1—∞ Payslip
/v1/attendance/check-ins
Record a mobile check-in with location and face checks
Request
{
"employee_id": "emp_01JD2K8V4R",
"site_id": "site_rimal_01",
"type": "in",
"captured_at": "2026-11-08T06:52:14+03:00",
"location": { "lat": 24.8384, "lng": 46.8627, "accuracy_m": 6 },
"selfie": "upl_9c41e2",
"device": { "id": "dev_7a2f", "mock_location": false, "rooted": false }
}
Response
{
"id": "chk_01JD2M3QZ8",
"status": "accepted",
"inside_geofence": true,
"distance_m": 38,
"liveness": { "passed": true, "score": 0.97 },
"face_match": 0.94,
"shift": { "code": "MORNING", "starts_at": "07:00", "late_by_min": 0 }
}
Webhook event · attendance.exception_raised
{
"event": "attendance.exception_raised",
"employee": "EMP-0644",
"site": "site_rimal_01",
"reason": "outside_geofence",
"outside_by_m": 140,
"notify": ["supervisor", "whatsapp"],
"at": "2026-11-08T07:04:51+03:00"
}
-
Integrations
-
Mudad
WPS payroll file upload, status and compliance rate
-
GOSI
Employee registration, wage changes and contributions
-
Qiwa
Contract authentication, work permits and Nitaqat data
-
Muqeem
Iqama and exit re-entry visa dates
-
Nafath
Employee sign-in with the national SSO
-
WhatsApp Business
AI assistant, request and approval notifications
-
Bank payroll
Salary transfers by bank file and IBAN validation
-
ERP
Posts the payroll journal to the accounting system
-
Security & compliance
- Data, photos and backups hosted inside the Kingdom.
- National Cybersecurity Authority Essential Cybersecurity Controls (ECC).
- Personal Data Protection Law (PDPL) alignment, treating face data as sensitive, with explicit consent and a fixed retention period.
- Selfies are deleted after the audit window, leaving only an encrypted face template; mock locations and rooted devices are rejected.
- Role- and department-scoped access; supervisors see team salaries only when granted.
- Maker-checker on payroll approval and on any change to a wage or IBAN.
- Immutable audit trail for every check-in, edit and approval.
-
Delivery plan
Clear phases with measurable deliverables.
-
01
Discovery & policy mapping
- Workshops with HR, payroll and site supervisors
- Inventory of policy, allowances and approval chains
2 weeks -
02
UX & architecture
- Employee app prototypes tested on a field site
- Payroll engine and data model design
3 weeks -
03
MVP build
- Geofenced attendance, shifts and leave
- Self-service and the HR dashboard
8 weeks -
04
Payroll & integrations
- Payroll engine and end of service
- Mudad, GOSI, Qiwa, Muqeem and the bank
5 weeks -
05
Assistant & parallel run
- Load and test the assistant on policy and the Labour Law
- A full month’s payroll run in parallel with the current system
3 weeks -
06
Go-live & hypercare
- Phased launch by site
2 weeks
- Total duration
- 23 weeks
- Team
- Project manager, an HR and payroll business analyst, 3 backend and frontend engineers, a mobile engineer, an ML engineer, QA and DevOps.
Performance
- < 5s Check-in including the face check
- < 48h Payroll closed after attendance cut-off
- 98%+ Mudad file accepted on first upload
- 70%+ Questions answered without HR hand-off
