Joe Philleo, Founder & CEO at Edia – Interview Series

Joe Philleo, the co-founder and CEO of Edia, leads the development of an AI-powered math platform designed for teachers and students, aiming to improve outcomes on state exams. The platform operates on the belief that education plays a crucial role in shaping individuals’ life trajectories. Edia’s mission is to create technology that ensures every student has access to an exceptional educational experience.

Currently, Edia collaborates with over 100 school districts across the United States, including prominent ones such as Fulton County, Loudoun County, and Palm Beach. These partnerships have demonstrated measurable success, with annual improvements in state math exam performance ranging from less than 2% to as much as 5-12%.

You made the bold decision to drop out of USC to work with Joe Lonsdale at 8VC, gaining exposure to groundbreaking tech projects. What were some of these projects?

Working with Joe Lonsdale at 8VC was an incredible opportunity. I left school when I was 20 years old to join his team, and it was my first exposure to Silicon Valley. Working in venture capital feels like living in the future—I met a lot of very smart people who were building self-driving cars, AI doctors, VR glasses, and new software systems to dramatically improve industry.

I spent a lot of time focused on defense, government, and education. I learned a lot. My biggest lesson from 8VC was the demystification of Silicon Valley. I grew up in Indiana, far away from any of this stuff. But spending time with Joe Lonsdale and other great entrepreneurs and investors made me understand that I could also make progress on solving big problems.

You’ve mentioned feeling disappointed that few top Silicon Valley teams were focused on K-12 education, which led you to start Edia in 2020. What specifically motivated you to tackle this gap in the education sector, and why did you feel the timing was right to launch Edia?

School has always been a personal obsession for me. I had three incredible teachers growing up who changed the trajectory of my life, and I also had some very bad experiences with teachers that pulled me in the other direction. Early on, I experimented with different ideas for how we could improve school. In 11th grade, I made a website called “booksarelong.com” to crowdsource AP textbook notes, and in college my friend and I applied Google’s PageRank algorithm to Wikipedia to build microcourses for all of human knowledge.

The real turning point came in 2020. Before then, only 10% of students in the United States had their own school-issued device, which severely limited how teachers and schools could use technology in their classrooms. Then, almost overnight, we went from 10% to 90% of students having devices because of the Covid-19 pandemic. In June 2020, OpenAI launched GPT-3, and it was clear that this was the moment to build something that could shape K-12 education.

What were the initial challenges you and your team faced when building Edia? How did you overcome them?

From the beginning, our team has been led by incredible engineers and designers. So, building our breakthrough math product wasn’t easy, but it also wasn’t the hardest part. As outsiders to the space, it took us a long time to distinguish our breakthrough AI math solution from legacy products that made big claims but rarely delivered. We felt that we had to distinguish Edia by showing real impact, which led us to guarantee growth for districts that use Edia—i.e. if they don’t see measurable results within one year of implementing, we offer a full refund. That promise has been a game-changer for building trust.

Can you explain how Edia’s AI math coaching works to provide real-time, personalized feedback for students?

One of the surprising things we realized early on was how little progress had been made in math learning software. Even fundamental challenges like “how do you easily do math on a computer keyboard?” hadn’t been solved before. We invented a completely new way for students to type math, inspired by Pinyin—the method Chinese and Japanese speakers use to type thousands of characters on a keyboard. This innovation makes it easier for students to show their work on a computer than on paper. Once the work is digital, AI can analyze it to understand the student’s thinking, identify where they went wrong, and deliver personalized inline feedback. The learning experience adapts to each student’s needs in real-time.

How does your platform use data to support teachers with small group instruction and data-driven lesson planning?

Small group instruction is one of the most impactful strategies for secondary math teachers, but it’s also one of the hardest to execute. Teachers often have 120 students across multiple classes and subjects, and it’s nearly impossible to pinpoint every student’s gaps, band them together, and create personalized lessons for each group. That’s where Edia comes in. Our platform automatically collects data from classroom assignments, quizzes, and homework to map out exactly where each student is struggling. Then Edia automatically forms small groups and generates customized lesson plans and practice tailored to their needs. This makes small group instruction manageable for teachers and has a huge impact on student outcomes.

Chronic absenteeism is a significant challenge in many districts—how does Edia’s platform address this issue uniquely with AI?

Chronic absenteeism—defined as missing 10% of school or more—has doubled since 2020, and it’s one of the biggest challenges districts face today. The key to solving chronic absenteeism is twofold: districts need to (1) understand and address why students are missing school and (2) rebuild the community’s expectation that coming to class matters. Our platform uses AI to engage parents within minutes of a student missing class to ask why their child is absent.

This interaction reinforces the importance of attendance, and it helps administrators understand the root causes of absenteeism—whether it’s issues with a specific teacher, social anxiety, lack of transportation, or something else. Armed with this information, schools can take meaningful action to address the problem.

Could you tell us more about the AI-driven, multilingual communication system and how it helps to engage families in real-time?

Engaging families effectively requires breaking down communication barriers. Many schools struggle to connect with parents who speak different languages or don’t check traditional forms of communication. Our AI-driven platform tackles this by sending real-time messages in the family’s preferred language, using conversational AI to bridge the gap. For example, if a student misses class, the system immediately reaches out to that family in Spanish, Chinese, Arabic, or any other language to let the family know and ask for an explanation – and parents can easily just reply back. It ensures families stay informed and engaged, while also helping schools address issues proactively. It’s about creating a two-way dialogue that fosters trust and accountability.

What is your long-term vision for Edia? How do you see the platform evolving in the next few years?

Our mission is for every student to have an exceptional experience in school. We want kids to look back 20 years after graduating and think, “Wow, I was really lucky. I had such a great time in school.” That’s the kind of lasting impact we’re aiming for. Right now, we’re focused on solving challenges that block students from succeeding. We began by making math accessible for everyone, and we’ve been able to accelerate growth on state exams from 0-2% per year to 8-23% per year, which is incredible. This year, we launched our AI solution to address chronic absenteeism, which is another critical barrier for many students to succeed.

But the possibilities are endless. From improving school budgeting to rethinking the design of school buildings—why do so many schools look like prisons?—to tackling critical issues like school safety, there’s so much to be done. Schools should be places that inspire, support, and protect kids. We see it as Edia’s mission to take on these challenges and ensure schools provide the best environments for growth and learning.

How do you see AI shaping the future of K-12 education?

One of the most exciting possibilities with AI is solving what’s known as Bloom’s Two Sigma Problem. In 1984, Benjamin Bloom found that replacing classroom instruction with one-on-one tutoring could improve student performance by two standard deviations, bringing an average student to the top of their class. But the problem is scale: there are 60 million students in the U.S. and only about 3 million staff. We simply don’t have enough adults to provide personalized tutoring for every child.

That’s where AI comes in. With AI, we have a real chance to give every student the benefits of a tutor. AI can scale this kind of personalized instruction in ways that were never possible before, helping every child reach their full potential.

What is the most rewarding part of your work at Edia, and how does it align with your personal mission in education?

It’s seeing the impact we’re having in such a diverse range of schools and students. We work with large urban districts like Fulton County in Georgia, New York City, and Palm Beach in Florida. At the same time, we’re also helping some of the smallest, most remote schools in northern Alaska, where the only way to get there is by seaplane or boat.

Knowing that students anywhere in America—whether in the heart of New York City or in a tiny Alaskan village—are both getting access to the same cutting-edge AI technology is really fulfilling. We’re giving these kids the experiences, confidence, and support they need to reach their goals, no matter where they are or what their circumstances might be. It’s an incredible privilege to play even a small role in shaping their futures.

Thank you for the great interview, readers who wish to learn more should visit Edia.

Joe Philleo, Founder & CEO at Edia – Interview Series

Related articles

Introductory time-series forecasting with torch

This is the first post in a series introducing time-series forecasting with torch. It does assume some prior...

Does GPT-4 Pass the Turing Test?

Large language models (LLMs) such as GPT-4 are considered technological marvels capable of passing the Turing test successfully....