Is AI the future of test prep?

This article is part of Reacheda series about companies that are harnessing new science and technology to solve problems in their industries.

Ever since Socrates taught Plato and Plato taught Aristotle, mankind has known that the best education is given individually by an experienced educator. But it’s expensive, labor-intensive, and difficult to scale. The result is the flawed classroom teaching we live with today: oversized classrooms, overworked and overworked teachers, a lack of resources. Educators focus what little time they have for personal attention either on the best and the brightest or on the bottom of the class. The wide middle is often left to its own devices.

Educators may have a new tool, AI, to solve these problems. Innovative forms of the technology, based on computer code that mimics neural networks in the human brain, can uncover patterns in student performance and can help teachers adjust their strategies accordingly. “AI tutors” — software systems that students interact with online — promise to give each student individualized attention, potentially remaking education as we know it.

Among the handful of companies driving this transformation is Riiid (pronounced “rid”), a start-up founded in Korea by YJ Jang, a graduate of the Haas School of Business at the University of California at Berkeley. Riiid already has a strong presence in the Asian market for test preparation apps for the Test of English for International Communication, or TOEIC, which measures proficiency in English for business. Now, Riiid is about to enter the SAT and ACT prep market in the United States.

“Education is a complex field deeply related to cognition, motivation, interaction with peers, etc.,” Jang wrote in an email. “We draw insights from learning science, cognitive biology, data science and other related research fields for an iterative process of experimentation that is difficult and time consuming – that is why there are only a few players in the market.”

The first computer tutoring systems appeared in the 1960s, presenting material in short segments, asking students questions and providing immediate feedback on the answers. Because these systems were expensive and computers were far from ubiquitous, research institutes were the main beneficiaries.

In the 1970s and 1980s, systems began to use rule-based artificial intelligence and cognitive theory. These approaches took students through every step of a problem, giving them guidance from expert knowledge bases. But rules-based systems failed because they weren’t scalable — and programming deep domain expertise was expensive and time-consuming.

Mr. Jang was evaluating such systems at Berkeley when a friend, Kangwook Lee, now a professor at the University of Wisconsin-Madison, introduced him to deep learning, a much more powerful form of AI in which algorithms learn on their own, drawing on mountains of data. Jang understood that deep learning could be applied to teaching, with systems that learn content and student behavior over time.

He returned to Korea and founded Riiid in 2014, working with a team of data scientists to develop a set of AI algorithms that track student performance, predict scores, and anticipate when students lose interest. and are about to give up. The company has published papers on this work at some of the world’s leading machine learning conferences.

To validate its technology and collect the data needed to refine its algorithms, Riiid launched a TOEIC test prep app called Santa (Santa Claus, of course, collects data on children around the world). It quickly became one of the best-selling educational apps in Japan and Korea.

Through the app, Riiid accumulated data on student interactions, creating what is now one of the largest public education datasets in the world, called EdNet. But Riiid struggled to collect enough data to generalize its AI system to the broader field of education.

“It is difficult to collect AI-trainable multimodal data in various learning environments,” Jang wrote.

For now, the company is focused on the $300 billion test prep market, where data is easier to collect, and has partnered with education companies in various regions around the world to develop test prep apps. Earlier this year, Riiid worked with Casa Grande to launch an app called OE Saber, which helps Colombian students prepare for the country’s Saber 11 university entrance exam.

Riiid’s success attracted a $175 million investment from venture capital giant SoftBank’s Vision Fund II, bringing the company’s funding to around $250 million.

Now, Riiid introduces an AI-powered preparation platform for SAT and ACT college entrance exams. The product, R.test, which will launch in January (pricing has yet to be announced), predicts standardized exam results in a quarter of the time it takes to complete a full mock test. By answering 30 questions, students get an analysis of their weaknesses and advice on how to improve, including an AI-curated selection of relevant practice questions. Riiid says the intent is for students to familiarize themselves with the app and take the real exam with some confidence in their final grades.

“I really liked it because we can just use it at home instead of hiring a tutor,” said Esther Yi, a parent in Georgia who tried an early version of the platform. She found the R.test analysis particularly powerful. “My 10th grader will definitely benefit from this,” she said.

Oscar Torres, a Chicago high school math teacher who tried Riiid’s system, said he liked R.test because it assesses students’ knowledge in real time without relying on past test scores. “As AI develops, I see it becoming a better and stronger resource for us,” he said. “We need to solve problems in real time as teachers, and AI can help us immensely.”

But the company’s focus remains broader than test preparation. Mr. Jang said R.test is part of an effort to collect data and prove the effectiveness of its algorithms in other areas. Riiid researchers continue to develop new architectures and better AI models that can track student behavior, trace student knowledge, and select the best content to study at any time.

“Our algorithms can predict students’ test scores with surprising accuracy, a moving number that serves as a carrot of sorts,” he said. “The more students follow the algorithmic recommendations, the higher their predicted score will increase.”

Mr. Jang believes that soon teaching will no longer be based on guesswork or hunches, but on data. And that’s perhaps the company’s biggest challenge: Collecting that data, he added, is the bottleneck because privacy concerns make data collection in schools complicated. (Riiid says its apps don’t collect any personally identifiable information from users.)

To address these concerns, Riiid helped establish the EdSAFE AI Alliance, a global, cross-industry alliance of companies, nonprofits, and edtech associations to develop benchmarks. and standards to ensure the safe and responsible use of AI in education.

“The dream,” Mr. Jang said, “is to integrate these algorithms into a complete system that can teach any subject to anyone, anywhere.”