This week, after many months of hard work and debugging, we launched the Faculty of Business and Law student support chatBot. The concept of a student support chatBot has been in development for many months, the idea been to answer some of the most common student concerns twenty-four hours a day, every day of the year, and without the need for a human ‘operator’.
Within that development time, various institutions have been looking at bots for a range of purposes, from supporting clearing to trouble shooting enquiries and more. JISC have sent a University-wide call for information on how various institutions are developing their use of such technologies, with a view to producing guidelines and best practice in the field.
The concept of our own chatBot started during conversations about the role of the Universities Academic Personal Tutors. The Faculty of Business and Law wanted to radically overhaul the system, and ran a three-day sprint to rapidly explore and develop solutions for a better, more student-centric approach to academic (and to some extent pastoral) support. You can watch a case study video from the sprint here. During these conversations, a mapping exercise was carried out to determine the number of places students could go for support. This exercise showcased the vast complexity of navigating the different support mechanisms available, and it was proposed that a central hub to triage information and speed up the process of finding a solution would be beneficial for both students and staff.
From a student point of view, it was seen that providing answers quickly and efficiently would enhance the student experience and troubleshoot problems and issues before they got out of hand. For staff, the reasoning was that by providing a digital assistant of some kind, academic time would be freed up to deal with more complex issues which need a human interaction to find a solution.
With the idea fresh in our minds, and some early research under our belt, we set out to design a bare-bones prototype to test the concept with the student body – using a Coventry.Domains hosting and LiveHelperChat open source software. We produced an artificial bot which was tested with the Faculties student representatives, with a user at the back end providing answers to the enquiries. Feedback from this proved overwhelmingly positive, with students commenting on the “great idea” of 24-hour student support.
Throughout our development of the bot, we were keen to minimise capital expenditure on the project as much as possible. Our first “working” prototype used a beta-platform developed by Quiriobot to begin structuring our developing FAQ bank into a key-word recognition system which could provide relevant answers to simple queries without the need of a human operator. Pleased with the usability of the software, support from the company, and the data compliance offered, we then progressed to the next stage of the development, opting to purchase a business license for 10,000 chats per month from the suppliers.
Click here to view a visual representation of the process of the bot.
In total, the capital expenditure for one year of running (assuming chat instances remain under 10,000 per month) is forecast at Euro 769.45. The build it fast, build it quick, and build it cheap mantra is core to the original intentions of the sprint, allowing us to test the offering with our students prior to an ‘all bells and whistles’ development. Throughout the process we have always pitched the offer as phase one development, and if student support is willing, we are keen to develop the bot further, possibly integrating into the University systems to provide a personalised level of support.
Currently, the bot is not a strategic IT project, instead designed, developed, tested, and owned by the Faculty of Business and Law and the Disruptive Media Learning Lab. Our hope in future is that the insights garnered from our piloting of a student support bot will fast-track the development of a University-wide platform, providing valuable data to enhance a fully supported platform.
For now, we’re going live with the pilot and awaiting feedback from students and staff as to the quality of interaction, the answers provided, and the general user experience. Watch this space.
For more information on this project please contact:
Paul Cashian
Associate Dean (Student Experience)
Faculty of Business and Law
Coventry University
Email: paul.cashian@coventry.ac.uk
Oliver Wood
Innovation and Community Producer
Disruptive Media Learning Lab
Coventry University
Email: oliver.wood@coventry.ac.uk
Header Photo by Craig Sybert on Unsplash