Case Study: Major UK healthcare provider reduces staffing costs by 11% and improves forecasting accuracy with Dataiku

How Dataiku has Enabled a Hospital to Significantly Decrease Staffing Costs and Turnover

Major Healthcare Providing Company, a UK healthcare provider of about 1,700 staff, was struggling with manual, bed‑based scheduling that led to physician overwork, patient dissatisfaction, and high staffing costs. To develop more accurate, transparent staffing decisions, they engaged Dataiku and its Data Science Studio (DSS) to build a data‑driven patient forecasting and staffing optimization capability.

Dataiku implemented an automated predictive application in DSS that combines internal and external data (historical volumes, weather, epidemics, holidays, traffic), trains ML models (ARIMA + decision trees), and exposes forecasts to the scheduling system via an API. The solution—delivered in under three months—made predictions 47% more accurate than historical averages, reduced staffing costs by 11% (about $730K/year), and cut estimated staffing turnover by ~9% in year one, improving productivity and care quality.