Soon after the pandemic inflicted by Covid-19 was declared, different organizations and smart application development companies suggested that artificial intelligence could be an useful tool in helping the medical staff monitoring the virus spread for a better understanding of its behavior. This paper presents a smart application CovidWatch developed for users real-time warning regarding the possibility of contracting the Sars-CoV-2 virus from an infected person. For this purpose, the developed application uses the Bluetooth module of the user’s device in order to check the social distance and to determine the possible contacts. The developed application has a number of four modules (the expert system (ES), the data retrieval, the real-time information and warning and the interface), that work on Android devices. Using the artificial intelligence techniques (knowledge-based systems), the application has a knowledge base populated with user’s data, the user’s time of contact and distance and the users identified as being contacts. If a user warns the android application that he has been infected with Covid-19, it further triggers the expert system inference engine, allowing the knowledge to be inferred. Based on the implemented rules, the inference engine analyses the data regarding the users that were in contact with an infected person. If those meet the conditions taken into consideration for declaring a person as being contact (a certain distance and time) they receive in real time a warning message and some advice regarding how it is recommended to act. The main advantage of android application is that it provides real-time warning regarding the possibility or probability of being infected with Covid-19, warning that can be deactivated only after the user has confirmed that the took note of this. A set of CovidWatch application simulation results is also presented taking into consideration different scenarios.