Strategic Education (owns Strayer and Capella Universities) is the US-based corporation focused on enabling economic mobility and seamless workforce adaptation to ever-changing skills in demand through affordable and tailored education.
The company targets mostly adult students that require special accommodation – e.g. they can take courses only in non-working hours – and often pursue customized degrees. These requirements coupled with the company’s mission to bring down the costs of education and provide the maximum ROI (both on time and money) for its students could only be met with a creative approach to the education model and Strategic Education has opted in for intensive use of AI.
First of all, Strategic Education has a vast set of complex but standardized data ranging from student admission files/polls to current academic performance and preferences. Having this resource obtained through the scale – almost 100,000 students enrolled, over 40 educational programs, and over 600 corporate employer partners – the company strives to employ predictive analytics for student behaviors to provide responsive learning and the most efficient and streamlined path to graduation. As an example, the company’s VP of AI cites the ability to “track missing assignments and use text interaction data to gauge behaviors, satisfaction, and learning style, crunching all of this information swiftly and effectively”. Currently this data is used only in automated interactions with the students, but Strategic Education plans to roll-out ML patterns recognition to adjust a student’s course schedule on the fly, gamify the experience, and provide instant feedback for professors – all aimed to facilitate the most effective learning experience.
Second, through the use of AI the company has significantly brought down the costs by eliminating unnecessary labor costs for academic advisors and IT helpdesk from its operating model. Strategic Education has already introduced Irving – virtual assistant – who is already addressing most of student questions ranging from financial aid to APA writing style, providing course recommendations and does that much more efficiently than human operator (e.g. average help session time was reduced by 3x). While almost 60,000 students have already interacted more than 350,000 times with Irving, there is ample room for further improvements. The current success rate (i.e. delivering the response expected by customer) is around 85% meaning that the rest 15% are redirected to human agents. By constantly training Irving on new sets of questions and extending its library of trained topics (currently 1,025) Strategic Education aims to do away with most of routine type of administrative tasks and/or often repeated questions being handled by humans. Further increasing Irving’s scope the company plans to move from reactive interactions with students to proactive advise like recommendations on joining a student group with aligned interests, reminding of the deadlines, and soliciting peer tutoring.
Finally, by recognizing the importance of tight AI integration in education process, Strategic Education should be well positioned to quickly grow its scale by both outcompeting other less flexible universities and creating a new market by providing good-enough education for obtaining new skills at a very low cost on the corporate tab. Given how quickly market needs for education are going to change in the future (what code syntax to learn? what marketing methods are going to be in demand?, etc.) an importance of keeping educational program up-to-date is only going to rise. Unlike traditional fundamental educational model employing courses tracing back to 1950s that change very slow, AI algorithms have the advantage of increasing speed-to-market of new solutions/courses and are much more easily scalable. Moreover, while traditional models use one-to-many format (i.e. all students listen to the same course) that works well for basic courses due to uniform demand for the knowledge the post-secondary education is focused much more on filing the knowledge/skills gap that are different for every individual. Therefore, AI-empowered way of matching unique student needs to the tailored curriculum (chosen from huge database curated by professors) is much more suitable for this many-to-one paradigm.
Strategic Education has an ambitious mission and has been so far only scratching the surface of AI capabilities in education. However, the company should move quickly by leveraging its scale and continue its fruitful partnerships with AI-enabling platforms (e.g. DialogFlow for Irving) to solidify its market leading positions. The costs of new tech-investments might be heavy, and the outcomes are risky, but I believe Strategic Education should make those bets by carefully experimenting in controlled environments (potentially involving volunteers or doing promotional test curriculums at zero cost). The potential first-order returns are quick growth and reduced costs while second-order benefits might include increasing caliber of alumni and professors choosing Strategic Education as a mean for knowledge transfer.