Multi-tenant Cloud SaaS Application for a meeting to task transition via deep learning models

Through digitalization and the post-pandemic periodonline meetings increase by 12.9% and the number of attendees by 13.5%. According to Zoom, Google Meets, and MS Teams, they collectively increased their amount of active users from March 2020 until June 2021 by 253%. Additionally, virtual meetings are becoming inefficient, unproductive, and unstructured, where results are not documented correctly. The intention is to shift those statistics by revolutionizing the way online meetings are held. The paper outlines the technical requirements for a user-friendly web application that allows reproducing a virtual meeting including agreements, its participant's interactions, and highlighting relevant sequences, is outlined. The developed AI meeting wizard autonomously documents the discussed content in the form of notes, to-dos, and tasks integrating them directly into task management, mail, or calendar tools. Reinforcing the focus on the discussion by increased interaction and productivity instead of task management and note-taking. Index Terms: Meeting Transcription, Deep Learning, Meeting to Action, Meeting Speech Recognition, Speech to Text