Introduction: The COVID-19 pandemic has posed great challenges in inpatient rehabilitation services, not only to implement the preventative measures to avoid the spreading of the virus in a highly interactive, multidisciplinary setting but also to create a rehabilitation pathway for post-COVID-19 patients. The aim of this retrospective study was to describe the role of a digital and artificial intelligence platform (DAIP) in facilitating the implementation of changes in a rehabilitation service during the COVID-19 pandemic. Materials and Methods: We gathered qualitative and quantitative descriptors of the DAIP, including measures to assess its efficiency in scheduling therapy sessions, and staff satisfaction using two simple numeric rating scales and the System Usability Scale. We describe how the volume of activity and the quality of care of our rehabilitation service have changed when the DAIP was implemented by comparing the pre-COVID-19 and the pandemic periods for patients' [sex, age, co-morbidities, diagnosis, and Functional Independence Measure (FIM) gain] and service's (bed occupancy, patients' length of stay, and staff capacity) characteristics. Results: Bed occupancy and the impact of rehabilitation on patients' outcome remained stable between the two periods. The DAIP provided a qualitative support for goal setting from remote; 95% of the planned sessions were delivered; the time for scheduling and registering sessions dropped by 50%. Staff satisfaction was about 70% for the easiness and 60% for the usefulness, and the mean "usability" score was close to the cut off for sufficient usability (mean score 65 where 68 is the cut off). Conclusion: By applying the DAIP to rehabilitation treatment, it was shown that the management of rehabilitation can be efficiently performed even in the COVID-19 pandemic. Staff satisfaction reflected a good acceptance of the changes considering the turbulent changes and the stress burden occurring at the time of the pandemic.