Client type
A services company that deploys employees to many client work sites. Each site has its own staffing requirement by role and shift, and employees have different skills, availability and locations. The client asked us not to publish their name.
The challenge
Allocation was done by supervisors in spreadsheets and over the phone. Every morning someone worked out who should go where, then called people one by one. This created familiar problems:
- Time lost every day on manual planning that followed the same rules each time.
- Gaps and double bookings when a change on one site wasn’t reflected elsewhere.
- No single view of which employee was at which site, on which shift, on any given date.
- Skill mismatches, where an employee without the required skill or clearance was sent to a site.
- Painful reporting for attendance, client billing and payroll, because data lived in several sheets.
What we built
We built a web-based employee management system with an allocation engine at its core.
Master data that reflects reality. Employees with roles, skills, certifications, home location and availability. Sites with required headcount per role and shift, plus any site-specific requirements.
Automatic allocation. The engine generates a plan for the selected period by matching site requirements to available employees. It respects hard rules (skills, availability, leave, maximum shifts) and then applies preferences such as keeping people on familiar sites and reducing travel. Supervisors review the proposed plan and adjust it before publishing.
Change handling. When someone calls in sick or a site changes its requirement, the system highlights the gap and suggests replacements that satisfy the same rules, instead of starting from scratch.
Visibility and reporting. A live board shows who is where. Attendance, site-wise deployment and employee-wise history reports are generated from the same data, which supports client billing and payroll preparation.
Access control. Admins, supervisors and managers see and change only what their role allows, with an audit trail on allocation changes.
Tech stack
Python (FastAPI) back end with PostgreSQL, a rule-based allocation engine with constraint checks, a responsive web front end that supervisors use on their phones, scheduled jobs for plan generation and notifications, and role-based access control.
Outcome
Allocation moved from a daily manual exercise to a reviewed, system-generated plan. Supervisors now spend their time handling exceptions instead of building the roster from zero. The business has one reliable record of who worked where, which makes billing and payroll preparation more straightforward. We describe outcomes qualitatively here because the client’s operational figures are confidential.
Could this work for you?
If your team assigns people, vehicles or equipment to sites, jobs or shifts using spreadsheets and phone calls, the same approach applies. See our custom software development service, or read how we handled a related problem in the multi-branch service management case study.