Frequently Asked Questions for Faculty & Staff About SAVY

What exactly is a virtual assistant?

The virtual assistant is a software-based agent that is populated with content designed on the IBM Watson platform to help you perform tasks related to your academic journey at York. It can either respond directly to your questions or point you toward accurate information relating to the questions you ask, recommend relevant resources, provide directions and wayfinding, remind you about upcoming deadlines and connect you to other resources, people and places on campus.

What is the difference between a virtual assistant and a chatbot?

While chatbots essentially feed you information by passing you content from other sources (like websites), a virtual assistant knows enough about you to provide you with more personalized content including reminders about upcoming deadlines and tasks relating specifically to you.

How is the virtual assistant different from just doing a Google search?

The virtual assistant is intended to give a single authoritative response to any request. Google offers many (sometimes thousands of) results for any search. In fact, Google will almost never generate a single result for a search because it will always try to use synonyms and other techniques to increase the number of results.

How do we differentiate the virtual assistant from all of the other informational resources for students?

From a technical perspective, the virtual assistant has the capacity to incorporate information from most existing current resources (websites, chatbots, etc.) and the hope is that the virtual assistant eventually becomes the single point of reference for information for York students.

Will the virtual assistant replace face-to-face student services?

No. The virtual assistant is meant to provide general information, even when it’s specific to your program. Human student service experts, like staff in Registrarial Services and advising centres, will continue to provide you with help for sensitive or complex issues requiring more in-depth support and conversations. Eventually, we hope that SAVY will be able to connect students to student service staff for in-person or live chat support.

Who can use SAVY, the virtual assistant?

Since SAVY, the virtual assistant, is embedded in eClass (formerly Moodle for Keele students), all undergraduate students will eventually have access to it. In the January 2020 release, students from the School of Arts, Media, Performance & Design, Glendon campus and the Lassonde School of Engineering gained access.

In the May 2020 release, students in Education, Environmental Studies, Health, Liberal Arts & Professional Studies and Science will have access. Students in the Schulich School of Business and Osgoode Hall Law School will have access in Fall 2020.

In September 2020, students in the Schulich School of Business received access, and SAVY launched in early October 2020 to Osgoode Hall Law School students.

Can graduate students use the virtual assistant?

No, the virtual assistant is only for undergraduate students.

How does the virtual assistant get its answers?

Subject matter experts across the University have contributed content to the virtual assistant to get started. During a test phase in 2019 students worked with the virtual assistant to help refine and expand content, and it continues to be constantly refreshed. As time goes on, the more it is used the better and more targeted the answers will get.

How will we develop additional content to support the system?

Subject matter experts and content creators are responsible for providing new and refreshed content based on continuous feedback from students using the tool.

How can I see what the students see on their Moodle page to better support them?

UIT is currently working on a providing access to the virtual assistant to specific student service staff, which should be complete in the coming weeks. In the meantime you can refer to this demo video for views of the virtual assistant in Moodle/eClass.

How will we generate and maintain student engagement with the tool?

In addition to word of mouth promotions, we will ensure that regular communication with students includes our key messages and clear, targeted calls to action. Communication to students will be across different channels with different ways to access the tool.

The naming contest during the three launch phases is also intended to generate excitement and engagement.

What happens if students don’t use the virtual assistant?

The project partners are working together to encourage and guide students’ use of the virtual assistant. Communication is aimed at ensuring that awareness and use of it is high and that its value is tangible.

How will the virtual assistant integrate with other pan-University initiatives (e.g. Web Optimization, SIS, CRM)?

Where relevant, the VA will pull data from SIS and the CRM to provide personalized answers.

How can we address the possibility that student service staff will be concerned about SAVY eventually replacing them?

We need to be aware of that concern and continue to emphasize the importance of face-to-face interactions. By providing general support to students, SAVY will actually enable student service staff to spend more time support students through complex issues. Eventually, we hope SAVY will be able to connect students to student service staff for in-person or live chat support.

How will we get staff buy-in to recommended using the virtual assistant to students?

Communications to staff will help them recognize the value of having the virtual assistant handle general queries, enabling them to provide more personal, caring and customized support to students on critical and sensitive issues.

How will the workload of content creators be managed?

University Information Technology (UIT) is building a content management system (CMS) to manage content workflow and streamline the content review, revision and approval process.

How will we ensure that departments aren’t duplicating each other’s content?

University Information Technology (UIT) is building a content management system (CMS) to manage content workflow and streamline the content review, revision and approval process. The CMS will enable an easy workflow, but human, subject-matter expert review of content will still be critical.

How will SAVY help administrative roadblocks to efficient student service?

With SAVY responding to general support and resource inquiries, we can expect a reduction in telephone and in-person inquiries, speeding up those response times.

Will SAVY change the way academic advisors interact with students?

No. Academic advisors will still be relied on to provide students with personal, customized academic guidance.

How can we ensure our competitive advantage once the virtual assistant is launched?

We are one of the first universities to launch a virtual assistant, but we can only maintain that competitive advantage through ongoing long-term institutional commitment to this system.

Can I get involved in developing content for or testing the virtual assistant?

We want your input and are currently creating a means by which you can provide feedback and suggestions for additional content. We will update this answer as soon as it’s available.

Where can I learn about York U’s other AI projects?

York is engaged with a number of research, teaching and learning activities in the area of Artificial Intelligence (AI), and you can learn more about them on the AI @YorkU website.

How will SAVY progress over time?

The more students use SAVY, the better it will get at providing the best possible answers to their questions. In addition to simply using it, students will also help train the virtual assistant by providing feedback through a feedback survey that is served to them whenever they say goodbye to SAVY. A Technical Issue Reporting Form is also available in eClass (formerly Moodle for Keele students) and the website.

Staff will review comments and feedback to continually see what should be added, improved or changed based on student need and impact.

Remember that SAVY isn’t human. We need to help train it to respond to queries.