While many discussions about the impact of AI on education have focused on what happens in the classroom, AI has the potential to address one of the ongoing challenges facing colleges, namely efficient customer service. Colleges need to be more effective in serving their customers, as evidenced by low student retention and completion rates. Of those who start college, 25% will not return for their second year, and just 60% will graduate within six years. The reasons most often given to justify the departure of students are financial: they come from inadequate support, whether administrative, financial or advice assistance.

College customer service teams face many challenges. Even though enrollment has declined, the number of applications received by many colleges has increased. Staff find themselves overwhelmed by a high volume of urgent requests, impacting response times and student satisfaction. Part of the challenge at colleges is that administration is often made up of students who do the work part-time and are eager to stop doing that work at the first opportunity. This leads to high turnover and low institutional knowledge on the front line. The challenge is compounded by the fact that different administrative systems are often not integrated, meaning customer service representatives must learn multiple systems and move manually between them. Traditionally, supervisors also have little ability to monitor these interactions, further complicating quality control.

AI to the rescue

Anurag Dhingra, head of collaboration at Cisco, has a bold vision for how AI can address these challenges. It begins by reorienting perspectives. The goal of introducing AI into customer service is not to replace humans with machines, but to equip people to be more capable of self-service. “Self-service is key here. Any time a user, customer, or customer needs or wants to accomplish something in a system but doesn’t know how to do it and resorts to calling support, it indicates a failure somewhere. This could be a failure of the product, product design, or business process design.

Designing a great self-service system requires understanding the user and what they are trying to accomplish. It also requires understanding the system and what needs to be done to achieve the various end goals. With these in hand, it is necessary to make available to the user the scope of things that would interest them and be capable of doing so. If there are too many choices, the system will lead to decision paralysis. If there are too few choices, the system will be frustrating. You need to find the Goldilocks median.

Faced with these challenges, Anurag Dhingra described how AI can make things better. To begin, the system may have a knowledge model of the user, including the context in which the user operates and their previous successes or frustrations. This should also distinguish between routine calls, for which AI may be preferred, and critical, emotionally charged calls, for which having a human on the phone is essential. Second, the system can place a much lower interface demand on the user because they can interact in ordinary language. This is true for both the user and the customer service representative, who may need to work on multiple systems that have not yet been integrated. Finally, thanks to AI translation, the system can move freely between languages.

The benefits of leveraging AI

A properly implemented AI customer service solution has the potential to transform not only the entire customer journey, but also the operation of the university bureaucracy itself. To begin with, the student experience is improved thanks to personalized and immediate support available 24/7. For universities offering programs for working adults or offering robust online programs in different geographies, this level of availability is essential. Additionally, when prospective students are early in the registration process, AI can be sensitive to context and the types of questions asked. It provides nuanced answers; if they are inadequate, the request can be forwarded to someone who can provide the answer.

Another benefit is the integration of digital messaging channels such as web messages and WhatsApp to further enhance the experience for students who prefer text over voice. AI can also be proactive in helping shy students, hesitant to ask someone questions, find the answers they need. After registration, the AI ​​can also look for patterns of behavior that indicate a student needs more direct human intervention, ideally before becoming a certificate hapless student.

Navigating bureaucratically complex systems should also improve. As the fiasco surrounding the Free Application for Federal Student Aid (FAFSA) has shown, the regulatory framework adds complexity that can be difficult for many students to understand. Replacing complex forms with dialog driven by AI agents should reduce frustration and improve response times.

As for the university itself, implementing a robust, AI-powered, self-service solution is expected to improve operational efficiency by freeing up more experienced staff to deal with more complex issues. Additionally, improving the interface and navigation between systems improves the worker experience, thereby reducing turnover. The system also helps collect data so the university can better understand its operations. This includes both user trajectory data as well as user satisfaction data. Finally, the ability to move freely between languages ​​addresses a problem that will become increasingly common in coming years, as more universities turn to international students to fill enrollment gaps caused by the demographic cliff.

Overall, AI has the potential to address critical challenges facing educational institutions. By adopting these solutions, institutions can improve communication, enhance the student journey, and streamline operations, ultimately leading to better educational experiences and outcomes.