Healthcare scheduling is a constraint satisfaction problem that people are usually asked to solve in a spreadsheet.
Every shift carries requirements: a qualification, sometimes a specific certification, a location, a time. Every person carries constraints: availability, contracted hours, statutory rest between shifts, travel distance, and preferences that quietly determine whether they stay with the employer. Past a certain size the number of valid combinations is far beyond what manual rostering can search.
What happens then is predictable. Coordinators stop optimising and start firefighting. Shifts get filled in order of urgency rather than in the order that produces good coverage, the same reliable people get asked repeatedly until they burn out, and one late cancellation triggers a chain of phone calls.
The cost shows up in two places: unfilled shifts, and turnover among the staff who absorbed the badly distributed work.
What the software has to get right
Hard and soft constraints are different things
Hard constraints cannot be violated: qualification requirements, statutory rest periods, contracted maximums, mandatory certifications. Soft constraints should be optimised but can bend: preferred locations, shift-pattern continuity, fair distribution of unsociable hours, keeping teams together.
Encoding that difference explicitly is what allows a system to generate a roster automatically without generating an illegal one. Systems that treat every rule the same either produce rosters that breach compliance or refuse to produce one at all.
Allocation is a solver, not a loop
Assigning shifts greedily in sequence produces a schedule that is valid at the start and impossible at the end, because early assignments consume the availability later ones needed. Allocation has to consider the period as a whole. That is the difference between filling most shifts and filling nearly all of them.
Fairness has to be measured to exist
Distributing nights, weekends and holidays evenly is not a nice-to-have; it is the constraint that determines retention. That means tracking who has carried what over time, not just within the current period, and treating the imbalance as an input to the next allocation.
Change is the normal state
The roster begins changing the moment it is published: sickness, cancellations, urgent cover. Re-allocation has to handle a mid-period change without rebuilding the entire schedule and without cascading disruption through people whose shifts were already confirmed and planned around.
Compliance is continuous, not a report
Rest periods, working time limits and qualification currency are checked as the roster is built, not audited afterwards. A system that lets you create a non-compliant roster and tells you later has moved the problem rather than solved it.
Every assignment records why
When somebody asks why a shift went to one person rather than another, and in staffing somebody always does, the answer should be a record rather than a reconstruction.
Where it fits
We build this as custom software where an off-the-shelf product constrains how the organisation actually works, which in healthcare staffing is common: qualification hierarchies, client-specific agreements and local regulation vary enough that configuration alone does not reach.
Where an existing patient administration or HR system is staying, the scheduling platform integrates with it rather than replacing it.