An approach to automating resource management of a service organization based on simulation modeling integrating the predicted input flows of patients, is considered. Methods for predicting input flows based on SARIMA, Holt-Winters, LSTM, and controlled recurrent GRU models have been investigated. The results of a computational experiment on predicting patient flows in a medical organization are presented. Based on the results, a meta-algorithm for forecasting the input flow and its further integration into the simulation model of the service process of a multidisciplinary healthcare organization was developed.