One of the key challenges in adopting a MBSE approach in industry is how to intuitively exploit system engineering knowledge expressed through descriptive models by different stakeholders. This talk presents a virtual assistant named IBIS in assisting system engineers to perform such activities over SysML models and data in multi-modal manner through text, speech or VR. Development of an underlying system engineering ontology, SysML model transformation, training a domain specific language model and query techniques are presented along with its implementation in the RASA framework based toolchain. Results of two different case studies are briefly discussed followed by lessons learnt, challenges and possible avenues of future work.