In CSCW immersive Virtual Reality environments, users may be uncomfortable when interacting with a two-dimensional menu. Multimodal conversational interfaces may enhance the interaction enabling users to communicate with the system in different modalities. In this paper, we investigate the use of an embodied multimodal chatbot for improving interaction in a Virtual Reality (VR) environment simulating a working context. In particular, we adopt a User-Centered Design approach to build a multimodal conversational interface, named Muxi, in which a task-oriented voice avatar is enhanced with an interactive board for supporting meeting organization in VR. Users were involved in all the development phases, from task definition to iterative user testing. To assess the usability of the proposed interface, we conducted a controlled experiment involving 32 participants to compare the use of Muxi with a traditional menu-based interface in a CSCW environment. We performed quantitative analysis, concerning efficiency and effectiveness assessment, and qualitative analysis, related to participant cognitive load and perceived usability. Results revealed that our multimodal interface increases usability by greatly alleviating cognitive load and improving user performance, representing a good alternative to a menu-based interface.