Optimal Resource Allocation
A Data Science Solution to intelligently assign care providers to participants based on required care levels and operational constraints
Business Challenge
Our Australian client, a provider of disability care services, needed to optimize resource allocation for participants requiring varying levels of care (e.g., 1:1, 1:2, 1:4).
The challenge was to:
Allocate staff while respecting individual care plans and weekly hour constraints.
Adjust staffing based on time-of-day needs(e.g., lower ratios at night, higher during the day).
Ensure regulatory compliance and cost efficiency in roster planning.
Approach & Solution
We built a Mixed Integer Linear Programming (MILP) model to generate optimal staffing assignments across shifts.
Care-Level Based Roster Optimization: Captured participant-specific care requirements (1:1, 1:2, etc.), daily time windows, and total weekly hour constraints to dynamically generate feasible allocations.
Solver-Driven Allocation Engine: Developed a Python engine to solve the MILP model and output an allocation matrix showing the optimal staff-to-participant mapping for each shift.
Impact
Optimized Resource Usage: Reduced overstaffing and understaffing through intelligent, constraint-based planning.
Improved Compliance & Quality of Care: Ensured delivery of care aligned with participant plans and government-mandated ratios.
Faster Roster Planning: Significantly cut down manual effort in schedule creation using a solver-backed, repeatable model.