Online collaborative learning environments are becoming increasingly popular in higher education. Etutors need to supervise, guide students and look out for conflicts within the online environment to ensure a successful learning experience. Web-based platforms allow for interactive elements such as conversational agents to disencumber the e-tutor. Repeatable tasks, which do not require a human response, can be automatized by these systems. The aim of this study is to identify and synthesize the tasks an e-tutor has and to investigate the automatisation potential with conversational agents. Using a design science research approach a literature review is conducted, identifying 13 tasks. Subsequently, a matrix is established, contrasting the tasks with requirements for the use of conversational agents. Furthermore, a virtual tutor framework is developed, clarifying the agent type selection, the technical structure and components for a prototype development in an online collaborative learning environment.