Founder Story

Reflections on His Journey from Google to Kyron Learning: An Interview with Rajen Sheth

By Malvika Bhagwat, Partner and Head of Outcomes at Owl Ventures

Malvika: What inspired you to leave your 17-year career at Google and start Kyron Learning?

Rajen: The mission of creating equitable access to great education is what inspired me to leave Google and start Kyron Learning. I’m privileged to have come from a family that gave me so many educational opportunities, but so many talented students don’t have that same opportunity to shine and find their potential. I want to focus my attention on ensuring that as many students as possible have the opportunities that I had.

I’ve always been inspired by great teachers; I think they are amongst the greatest heroes in our society. Also, I think we are at an inflection point where technology, in particular AI, is evolving so quickly that it’s opening up potential for new solutions every few months. I want to thoughtfully and responsibly apply AI in a way that empowers and amplifies great teachers.

Malvika: What are the most important learnings and reflections you are bringing to Kyron Learning from your time at Google?

Rajen: My time at Google was a once in a lifetime experience to see a great company grow from a small company to what it is today, and I learned a lot from it. When I joined, Google was only about 2000 people, and it grew to over 150,000 employees by the time I left. There were three critical lessons I learned through my time there:

Always Put the User First

There were many times at Google where we could have put revenue above user experience, and we always chose to make the experience great for the user. This created a better product and actually ended up creating a better business as a result. I learned to build a hypothesis on what matters to the user, create the best experience for that user, and use their feedback to get better. With Kyron, our ultimate users are the student and the teacher, and we are hyper-focused on making sure that we build something that’s great for them.

Fail Fast and Iterate

We tried a lot of different things at Google, but we were also good at experimenting and letting the results shape our direction. I remember how, when we started Chromebooks for Education, we tried many different product ideas and paths to market. We even experimented with a Chromebook vending machine! Doing this allowed us to find the product and path which worked best for schools, and it ended up having a big impact. We learned from each and every failure along the way.

Be Bold

We often tried crazy ideas with a big vision behind them. Sometimes they didn’t work, but many times they did as we forced ourselves to think about how we could radically change the user’s experience for the better. This often took patience. When we started Google Apps for Enterprise and Education, I literally had customers walk out of the room five minutes into the presentation because they didn’t like the idea of hosting their email in another data center. With Chromebooks, an analyst even labeled it “corporate idiocy.” But, we kept pursuing bold visions and, often, it played out well. This is harder to do at a startup, but it is still something I think is important for any company

Malvika: You mentioned that your goal is to “supplement teacher[s] and provide them with a view of where they should target their instruction for each of their students.” Can you talk more about this? What value will the Kyron Learning platform bring to teachers and students on a day-to-day basis?

Rajen: Classroom teachers have a very hard job in general, and it has become even harder post-COVID. In the average math class, there’s a 3- to 4-grade variance in terms of skill level, and a teacher faces the nearly impossible task of trying to meet students where they are, while shepherding the whole class to its achievement goals.

We want to give teachers a tool that can help them meet students where they are. Imagine if a teacher was able to assess an area where a student needed help and then assign a lesson that feels like a tutoring session to the student. This tool could help teachers meet the gaps that students have and bring students back up to grade level. This is the key problem we are trying to solve with Kyron Learning.

Malvika: What do you anticipate being the benefits of working hand-in-hand with educators during the product development process?

Rajen: We have seen huge benefits of working hand-in-hand with educators during the product development process; we wouldn’t be able to build an effective product without them. I have to fully admit that I’m a technologist and not an educator, and I still have a lot to learn about education. I rely on working with educators to make sure Kyron Learning is doing something that truly helps them.

We have several experienced technologists on our team, and we also have several experienced educators who have taught in under-resourced communities. We have been able to share our respective learnings, and it’s made for a wonderful learning experience for everyone.

Our teachers are learning about the nuances of natural language understanding and model training, and our technologists are learning about the difference between conceptual understanding and procedural fluency in math. We have AI engineers literally working side-by-side with educators to create our first lessons, and it’s allowed us to iterate on both the technical design AND the educational process.

Moreover, we’ve established a council of education leaders and a council of math teachers. This has formed a group of people that we can turn to when we are stuck and need advice, either on the details of pedagogy or the broader picture of school district priorities.

"Kyron Learning should give these students access to incredible education that can supplement the great work of their teachers."

Rajen Sheth, Founder and CEO of Kyron Learning

Malvika: What’s your vision for where Kyron Learning will be in the coming years? How will you determine if you are making your intended positive impact on students and teachers?

Rajen: My vision is that, five years from now, we will have empowered thousands of teachers to share their talents with millions of students, and, most importantly, we will drive both enthusiasm and effective learning in math plus many other subjects. In particular, I want this impact to be focused on under-resourced communities that typically wouldn’t have access to high-quality tutoring. Kyron Learning should give these students access to incredible education that can supplement the great work of their teachers.

We would measure this impact with data, of course, looking at whether we have enabled students to love math and succeed in math where they weren’t before. But, I also want to follow the stories of students for whom we helped pave a path to a better future.

When I worked on Chromebooks for Education, we saw in the third year that many schools were starting to use the product, and adoption was growing. Yet, what really drove home our impact was stepping into a classroom at Urban Promise Academy in Oakland and seeing a 7th-grade student show us the game she was able to code because she had a Chromebook. It was the first time that she had access to a computer at home, and it formatively shaped her academic journey. I will always remember this moment; it was one of the highlights of my career. I am hoping we will create many such stories and opportunities for students through Kyron Learning.

Malvika: Finally, there has been so much talk about new AI technologies like ChatGPT, and the capabilities of AI are evolving quickly. How will this technology evolve over the next five to ten years, and how can this be applied to education?

Rajen: I feel like we are approaching a moment with AI that is similar to what happened when the internet browser was first released in 1993. When the browser was released, the internet had been around for nearly 25 years, but the browser all of a sudden gave everyone a view into the power of the internet. Similarly, AI technology has been around since the 1960s, but over the last decade, everyone has seen the power of how AI can impact their lives. ChatGPT just opened a big window to everyone into the most cutting edge AI technology called large language models (LLMs), and it is amazing to see how much awareness it has brought to this area that has been building up over the last 5-7 years.

This will evolve rapidly now over the next 5-10 years, and we will see it be able to answer many more questions much more accurately. We will also see a plethora of applications built on top of it. In education, this can help with everything from answering questions for students, creating content for educators, and becoming a new way to do research. However, it’s still at a very early stage, and the accuracy of response is still not where it needs to be for many uses, and often it’s not easy to figure out how it came up with that response. This will get better and better, but for critical use cases like education, it needs to get to 99.9% accuracy to put in front of a student.

As a result, there are three things we need to think about when applying this to education:

Empowering and Amplifying the Educator

The most successful implementations of AI are ones where the AI is making recommendations to an expert. For example, we still don’t have driverless cars widely used, but technologies that help a driver (adaptive cruise control, lane keeping, etc.) are now standard on most new cars. In education, the expert is the educator, and AI can be used to empower and amplify the educator. This is what we are trying to do with Kyron Learning, and we want to be a place where teachers can come to make use of AI to help their craft.

Educating the Students About this Technology

There will be more and more technologies out there which will give students a more exact answer to their questions. Students need to be trained to have a discerning eye to figure out whether the information they get is correct or not. Moreover, we need to think about how this can be a tool for students, but also how it will not get in the way of student mastery of a skill.

Understanding Responsible AI

AI is prone to bias based on the data that it is built upon, and it can have a dramatically positive and dramatically negative effect depending on how it’s used. As this technology is more widely used, we all need to understand how to use AI responsibly and create that positive effect.

We’re in the midst of a revolution in AI, and I’m happy that we’re at the forefront of using AI to amplify educators.

Want to keep reading?

Founder Story

Changing the App Development Landscape with No-Code: An Interview with Arun Saigal, Co-Founder and CEO of Thunkable

We at Owl believe that Thunkable is going to be one of the most powerful and ubiquitous tools in the world of technology, and we are thrilled to announce that Owl has led Thunkable’s Series B. Read more

Stay up to date on the latest information in from the leaders in Education Technology