Chatbots in universities
Today’s universities are challenged with pretty intense competition for the best and brightest students, while retaining and nurturing those students throughout their college careers. It’s important to make sure students are the right fit from the start, while easing their transition to college and providing as much support as possible–a pretty tall order under any circumstances.
Yet, what makes the challenge even more difficult is making sure universities have the resources to address each student’s questions and concerns while also becoming proactive in helping them maintain good grades, ensure financial aid eligibility and work to their greatest potential.
This is where chatbots and other forms of AI-based technology are making the grade. A chatbot is an Artificial Intelligence (AI)-based solution that conducts a conversation via voice or text, simulating how a human would behave in a conversation.
Chatbots enable universities to connect with students right where they are–and that’s usually online. Today’s chatbots can communicate with students on the university website, through Facebook Messenger, SMS or other mobile channels.
Send in the bots
The Inter-American University of Puerto Rico recently employed an online chatbot in its student services center to help field hundreds of questions around the clock, about financial aid eligibility, academic programs or even library hours during exam week. By freeing up service center staff, the department can now focus more on engaging with students on a higher level. They’re now able to help identify students in danger of failing grades, or those who may qualify for additional financial aid; and they can initiate outbound texts to let students know.
The chatbot is trained to answer 80 percent of the frequently asked questions and it’s being integrated to interface with other systems throughout the university, for example soon students will be able to receive student transcripts or the status of grants.
The possibilities of employing chatbots on university campuses are endless. Imagine using a chatbot in Admissions to handle the myriad of questions that come in during the admissions process; or even to interview candidates, engaging in dialogue with students and recording responses. Many universities today are eliminating the interview option for prospective students simply because there isn’t enough time, given the thousands of applications needing review, along with test scores, essays, etc. But the need to properly vet students and ensure they’re the right fit for the school is more important than ever. According to a recent article, one in three first-year students or more won’t make it back for sophomore year.
Other trends, such as the growth of online degrees, is driving interest in chatbots as well.
As students increasingly take online courses, chatbots can provide a valuable resource, becoming a counselor of sorts. Universities that offer both online and campus programs, can provide seamless integration of students wherever they may reside, using the chatbot as an intermediary. For example, chatbots can answer student questions about assignments, prompt students to meet deadlines and provide good advice about where to find extra help or additional resources on specific subjects.
Additionally, while curriculum, admissions deadlines, events and other information continuously changes, it’s not realistic or cost effective for universities to constantly update their websites. Chatbots allow them to address student questions that may not be available by surfing the website. Chatbots can be trained to dispense the latest news through a chat window, and this information can provide valuable insight for the university, which can hear about complaints, concerns and information that may not be available on the website.
In virtually every university department, there’s a role that chatbots can play, but what are some of the key considerations for universities looking to take the plunge?
Consider your objectives. It’s important to work across departments and identify what the business problems is you are trying to solve: Do you want to improve student retention, simplify the admissions process or make it easier to conduct remote learning? Once you determine the key business problems, you can more effectively refine the chatbot to meet them. It’s also important to think about how much time and focus you allot to solving the problem.
Apply the 80/20 rule. In this rule, roughly 80 percent of the effects come from 20 percent of the causes, is a good rule of thumb with chatbots. If you’re designing a chatbot to speak with incoming students, having it understand speech 80% of the time will probably do the trick. While it’s hard for academic-driven professionals to pull back from perfection, it is often the right thing to do from a business perspective.
Be picky about the platform. There are a variety of advanced tools, such as Microsoft Azure, Google TensorFlow and IBM Watson, that can help you build a complex chatbot affordably, and your data scientist can help you identify the best platform to meet your needs. Many of these platforms represent the future of AI, and because they provide a ready-made foundation, you don’t have to build a solution from scratch, freeing you up to focus on solving business problems.
Understand that it’s not one and done. It’s important to understand that, unlike other software apps, AI-based solutions are never really completed. While their benefits can be realized immediately, as smart solutions they need to be constantly trained with new information in order to stay on point with student and other university needs.
In an increasingly competitive higher education marketplace, where universities work to attract and retain the best and brightest students, chatbots are making the grade, helping to improve the student experience and provide the support, encouragement and resources within the environment in which students are most comfortable—the digital world.
Carlos Melendez is co-founder of Puerto Rico-based software engineering services firm, Wovenware