A look at how the company uses AI to improve hearing aids, what inspired the company, and moreRead more...
A look into how AI and chatbots can streamline higher education
Schools across the world are starting to adopt more technologies into their processes, but it is still a slow going. For one, you have to build on antiquated systems, but there are also issues with privacy and student information.
Educational institutions have to be wary of adoption when they store so much personal information about students, many of them younger than 18. Even so, there are companies working around the clock to bring the education system into the 21st century.
Take, for example, AdmitHub, which uses AI and chatbots to help guide students through some of the processes involved with higher education, while efficiently escalating certain issues to advisors and the proper departments.
I had the chance to speak with Drew Magliozzi, CEO and co-founder of AdmitHub, to learn more about the service and the technology behind. Check it out below.
Care to introduce yourself and your role with AdmitHub?
I'm Drew Magliozzi, AdmitHub’s CEO and co-founder. I've dedicated my career to helping students succeed. Previously, I founded a tutoring company and non-profit open education venture funded by the Hewlett Foundation. I graduated from Harvard in 2005 and Hack Reactor, an intensive software engineering program, in 2013.
My co-founder Kirk Daulerio is AdmitHub’s CMO. He was the first person in his family to go to college. He was both an athletic recruit and Pell Grant recipient. After graduating, he worked in admissions at Princeton, Penn, and Bowdoin to help others have a transformative college experience. He also served as Director of College Counseling at Gould Academy and later as Director of Member Relations for The Common Application. He holds a BA in Psychology from Swarthmore and an MSEd in Higher Ed Administration from the University of Pennsylvania.
We understand the tremendous value of higher education and care deeply about the education of future generations. We feel a responsibility to bring students, educators, and advisors together with our powerful communication platform for their mutual benefit. Many members of our team come from low income, minority, and first-generation college backgrounds, and it is this personal experience that drives our work. Our passion for helping young people attain the transformational life experience of a college education was the impetus for launching AdmitHub in 2014.
In just a couple sentences, what does AdmitHub do?
Broadly, AdmitHub fosters student success with mobile messaging powered by AI.
Specifically, we build chatbots that combine the best of behavioral economics and artificial intelligence to engage and support students on the path to and through college.
Colleges and universities are our customers and our bots are often branded as their school mascots to guide students through complex tasks, gather data, conduct surveys, involve advisors, and respond to their questions 24/7.
What problems are you trying to solve?
We’re trying to reduce the points of friction for students and administrators when it comes to admissions, guidance, and just being in college. Here’s one example:
In late 2016, we wrapped up a randomized control trial with our first institutional partner, Georgia State University. That project was focused on solving their “summer melt” problem, which was approaching 20% of committed students failing to enroll each fall. To help them solve this issue, AdmitHub built Pounce (GSU’s fuzzy panther mascot) and partnered with Dr. Lindsay Page (University of Pittsburgh) and Dr. Hunter Gehlbach (UCSB) to conduct a research study based on the gold standard of scientific rigor.
By engaging half of all admits (and holding the rest in a randomized control group), Pounce drove a 3.9% increase in overall enrollment and a 21.4% drop in summer melt.
While the end result of increased enrollment and decreased melt at GSU and other partners is exciting, the more interesting insight came from all of the intermediate tasks and processes. Specifically, AdmitHub has proven effective at moving the needle on every large group enrollment action we’ve confronted.
There is no silver bullet in college success. Instead, we focus on all the small nudges and positive interactions that add up to a significant impact in the end. We’re expanding into retention to provide personalized support in every facet of collegiate life, from financial aid re-application and scholarship matching to a sense of belonging, and eventually, internship/career advising.
How many schools are you currently in?
We have 25 partners (and counting) spanning the breadth of higher education in the US and abroad, from large public research universities to HBCUs to mid-sized privates to small liberal arts colleges to community colleges.
Are any more schools coming?
Yes! And the ones we have are often expanding their scope of work with us.
As a startup, what kind of hurdles did you face early on?
Selling artificial intelligence to an industry that tends to be conservative with regards to technology was quite a challenge at first. Luckily, a series of internal and external changes conspired to help us.
First, we learned how to talk to colleges about our technology. Hint, don’t mention convolutional neural nets. Second, the proof point of our research study at GSU was too impactful to ignore. And finally, artificial intelligence has become steadily more mainstream and even appealing to all industries.
Have all of the school systems been receptive to the idea of AdmitHub?
Usually, the first reaction is a mix of fear, apprehension, and hope. We like to say that our partners are not so much buying software, as hiring it. Like any new employee, there is a period of training and acclimation before one of our Mascot Bots is performing optimally. Our onboarding process tends to take eight to ten weeks and by the end, all of our partners are able to recognize the opportunity, personality, and scale that Pounce, Chompers, Sunny or any of our Mascot Bots is capable of with their support.
Anything exciting coming to the platform?
Lots but we’re not ready to share just yet. Please stay tuned!
The big question - how do you feel about the current state of education? What needs to change?
The biggest challenge in higher ed is data management. Right now, FERPA and GDPR have created a climate of fear around accessing and sharing data. This is a justified concern, but the result is that data (even within the institution) is siloed and inaccessible across departments and with companies like AdmitHub.
As a result of these endemic data silos, we have stifled innovation in the space. This has not been beneficial in an era of traditional computing, but it’s increasingly problematic in a world where AI is poised to make dramatic gains.
The crux, without delving too deeply, is that merely getting a data dump periodically is not enough for most predictive deep learning models. In order to work optimally, predictive models need real-time access to “time series” data, which creates huge switching costs and dramatically slows innovation.
We need to move toward a future where colleges can create interoperability and data-sharing without sacrificing security. The need is fast approaching and the technology to solve this challenge is here, but it requires a concerted effort to create a decentralized network for solving these issue. Our higher education institutions created the first internet. It’s time for them to self-organize for the creation, to quote Richard Hendricks, of the “new internet.”
How do students feel about AdmitHub?
- “It was the easiest part of enrollment”
- “It's useful to gauging and improving the student experience”
- “It's faster than any other assistance format”
- “It helps keep you on track and reminds you of important info needed and things you may have missed”
- “It makes the enrollment process more personal”
Some students prefer talking with their chatbot more than with a human because they “didn’t feel judged” by asking “simple” questions.
The most common student message is “thank you.”
Support VatorNews by Donating
Read more from our "Interviews" series
The digital clinical trial company raised $91M, bringing its total funding to over $136 millionRead more...
A look at Machine Learning Data Catalogs, how the company helps clients manage data, and moreRead more...