Resource Optimization for a Chicago-based Hospital Network



accuracy increase in forecast


FTE reduction of hospital staff

The Challenge

A large hospital network was struggling to effectively allocate resources to cater to the needs of incoming patients. They were faced with large emergency queue lines since they were either short staffed, or did not have access to the proper equipment to treat patients. They needed a way to predict the kinds of patients being admitted in their hospital over a certain period of time, as a way to optimize the use of the resources they had at hand.

The Solution

The client sought out Addo’s services to come up with an answer to their asset utilization issues. We developed a resource management solution using AI techniques to optimize our client’s hospital staffing problems. The developed AI solution was capable of predicting the number of new patient admissions, as well as the rooms or facilities the patients would most likely need to use once admitted into the hospital. Moreover, the predictions could be made for the next seven (7) days, and as a result, the forecasts generated were incorporated into our client’s weekly staff assignment process.

By collecting, analyzing and preprocessing resource occupancy data with the help of several AI techniques — which included time series and regression analyses — and by engaging a pool of highly skilled Machine Learning Engineers, we were able to provision a solution for our client.

AI techniques: Bayesian Time Series Analysis, Facebook Prophet, Lasso Regression, Vector Autoregressive Model

The Results

  • Increase in forecasting accuracy
  • Reduced costs
  • More efficient use of resources

Technologies Used

Jupyter Notebook

Looking for a
similar project?