Microsoft Teams and AI transform student experience

Doctor David Kellermann’s Microsoft Inspire 2019 presentation blew us all away! The story of how he transformed a Mechanical engineering class of “500 islands” into a single learning community, continues to motivate educational institutions to achieve more (and even enterprises including groups at Microsoft). We have since been eagerly waiting to leverage his unique and remarkable brainchild – the QBot! Today, we’re thrilled to announce that QBot – the learning community bot – is available as an opensource app on GitHub.

Microsoft Teams has fundamentally changed how students at UNSW learn, in and beyond the classroom. Previously, as is the norm with large classes where a tutor is responsible for 100+ students, important questions get missed, with each one representing a missed opportunity to make a lasting impact with a student. This led Dr. Kellermann to the design of QBot – an app built on the Microsoft Teams developer platform. The idea, started as a means to keep track of both answered and unanswered questions in a Teams class team, and to send notifications to designated class tutors when new questions appeared. Once a repository of Question/Answer pairs had been generated, QBot could then harness Azure’s Cognitive Services to learn, and proactively respond to incoming questions.

Over the last few months, Microsoft Teams product engineering has partnered with Antares Solutions, and the University of New South Wales’ Dr. Kellermann, to rearchitect and opensource QBot. Whilst Antares and Microsoft are working to implement the full solution in a number of universities already, the ‘lite’ opensourced version of QBot is now available to any educational institution in the world to deploy across multiple courses, and/or to enhance to better meet their organisation’s unique needs or pedagogical approaches.

The core experience of QBot remains just as familiar and intuitive in the opensource version. Anyone in the team can tag QBot in a channel and ask a question. This then prompts QBot to notify the right tutor(s) – so that no question goes unanswered. When the community responds with answers, the author or a tutor can choose the best response, which in turn lets QBot “learn” and build a knowledge database of high-quality question/answer pairs. Over time, QBot will instantly respond to new questions posted in the channel.

What’s more – as QBot keeps track of answered and unanswered questions in each channel as coinciding pinned dashboard, students and tutors can instantly access questions coming from their tutorial groups – right from within Microsoft Teams. Students and teachers can then effortlessly skim through, revise their understanding, respond to, or even prepare for the next class! Microsoft’s Azure powered knowledge base further provides lecturers and tutors with the ability to view trending questions, and to tailor responses to supervise QBot’s learning. Lecturers and tutors have additional aggregated views, so that they can view student engagement across their classes. Course administrators can also configure QBot from within the app.

As an Azure AI-based learning platform, QBot is designed to become part of the natural human interactions that happen within a community of learning. QBot connects students and faculty in a casual, communal, and approachable way – which promotes engagement, inclusion, and learning. Thanks to the hard work of Dr. Kellermann, the University of New South WalesAntares Solutions, and Microsoft all of this is now available for you to foster a collaborative learning community in your Teams environment.

Are you ready to get started? Everything you need is on the QBot page on Github

This story was sourced from Microsoft's Education Blog