Meet the 5 Education Technology Startups From Y Combinator’s Summer 2017 Class

Investor pitch events are a dime a dozen, but no one does it with the flair and furor of Y Combinator. Now in its 25th Demo Day, the Silicon Valley accelerator has created a machine for turning fledgling ideas into businesses that get investors salivating—and writing checks. Since 2005, Y Combinator has funded 1,430 companies and nearly 3,500 founders. Its graduates have gone on to raise more than $13 billion in total.

The latest cohort, unveiled over two days at the Computer History Museum in Mountain View, Calif., featured 124 companies from 18 countries—marking the largest class in the accelerator’s history. Of this group, five teams are building education technology tools with support from mentors from Imagine K12, the local edtech accelerator that Y Combinator absorbed last February.

Here’s what they’re working on:

Lambda School

The San Francisco-based startup combines some of the hottest trends in the edtech industry: computer science education, small classes, and income-share agreements. Lambda School allows students to pay nothing upfront, in return for 17 percent of their salary for two years once they get a software-related job that pays more than $50,000. (The company will not take more than $30,000 over the two years.)

Other tuition options exist: a student who pays $10,000 upfront will give up 17 percent of their income for one year. Those who pay $20,000 don’t have to give anything back.

“Both my co-founder and I come from small towns, where we were surrounded by brilliant people who don’t have the opportunities that we enjoyed,” Austin Allred, the company’s co-founder and CEO, tells EdSurge. “Our main goal is to give access to world-class technical education for folks who may not otherwise have it.” The company accepted its first cohort of 30 students this past July, added another 40 in August, and will have 100 for its September class.

The company currently offers two live, online and interactive 6-month programs in software engineering and machine learning. “Intensive” doesn’t quite capture the dedication that these courses require. “You’re going hard, 9 to 6, Monday to Friday,” Allred adds. Each program covers the basic programming skills for the first three months, followed by two months of computer science fundamentals, and wraps up with a one-month capstone project. The company says it has 50 hiring partners and plans to grow the list.

Mystery Science

Building lesson plans is not a track that many edtech startups venture on. And neither is elementary-grade science. But Mystery Science hopes to buck the trend. The San Francisco-based startup helps teachers deliver engaging science lessons through a series of videos, discussion questions and lesson plans—all packaged as a “mystery.” The company claims it has reached 1 million children in the U.S., with teachers using its materials on a monthly basis in 20 percent of elementary schools.

“The early years are the most important years to develop a curiosity about the world,” co-founder Keith Schacht told EdSurge in June, after closing a $2 million fundraise. “Elementary school is often a forgotten place but it’s the most critical.”

The company also has a knack for timely marketing, having recently partnered with Google to deliver 15,000 glasses—for free—to elementary school teachers in time for the eclipse. TechCrunch reports the company is forecasting $3.5 million in revenues this year and $1.5 million in profit.

Nimble

Hire faster, hire better, and reduce teacher turnover. Led by a former D.C. Public Schools human talent administrator, Nimble aims to be “a better applicant system that uses machine learning to find teachers with the best potential and match them in classrooms where they’re likely to succeed,” says Lauren Dachille, the company’s co-founder and CEO.

The traditional hiring process for teachers is disorganized, she claims, and often involves disparate, disconnected tools. “Sometimes districts are hiring for convenience, not quality.”

Dachille believes that data can determine the type of classroom environment where a teacher can thrive. Some of that information can come from résumés and work samples (in the form of videos, for instance). But the system also collects unstructured data through open-response questions, and uses natural language processing to parse those answers and identify where an applicant may be strongest.

She says the company has entered into some data-sharing agreements with large U.S. districts, and plans to use that anonymized, historical data to “build the foundation of our matching tools.” The company has tested the tool in nine school districts and charter management organizations, and plans to launch publicly in October.

Peergrade

Feedback is one of the essential ingredients to getting better—at school, work or life in general. It’s even better if your classmates or colleagues are dishing the tough love. That’s one of the founding philosophies behind Peergrade, a Danish startup is building a set of tools to make it easier for instructors to incorporate peer grading and feedback into the classroom.

“Peer feedback is an old concept,” David Wind, the company’s co-founder and CEO, tells EdSurge, “but it is inherently a complex process where lots of synchronization needs to happen.” Tools like Google Docs allow for real-time feedback to happen, but keeping track of all the comments in a classroom of 30—or a lecture hall of 300—can be a massive headache. One of the company’s newer features, Peergrade Live, allows educators to create and distribute assignments on which students can give feedback on, and track the quantity and quality of such comments.

The company claims users in every university in Denmark, its home country. (Half of them are paying customers, Wind adds.) It’s making headway into the K-12 market, which currently makes up roughly half of all its users. TechCrunch reports the company is already making $150,000 in annual recurring revenue.

Py

It’s that familiar story of Ivy League students who built something during their spare time: The duo behind Py created their first mobile course on Python as undergraduates at Yale and Brown, which proved popular in Russia. By the summer of 2016, they had six courses covering JavaScript, SQL and other typical languages. Today Py’s content library covers more than 22 courses for skills like Swift and data science.

“Lots of people would ask us: ‘How do we learn to code?’” recalls Derek Lo, Py’s co-founder and CEO. “We created this for people who don’t know how to code. We wanted to offer an easy entry point,” he adds. The team’s target audience: “the Snapchat generation.” The company boasts more than 100,000 users and 200,000 downloads, Lo shares.

Key to winning over a learner’s attention, he adds, is breaking down the courses into digestible chunks. Each course is made up of 70-odd short lessons and exercises. To monetize, the company plans on embarking on the familiar route of matching learners with employers. “Based on our data, we can rank users according to specific categories and skills and match them to job opportunities.”

This revenue will help the team “keep the basic educational courses free,” states Lo. “I cannot envision a world where our intro materials are blocked off behind a paywall.”