The Center for Data Innovation spoke with Varun Gulati, co-founder and CEO of Koalluh, a generative artificial intelligence (AI) platform that enables children to develop personalized stories for reading practice. Gulati spoke about the potential of generative AI to make classrooms more collaborative and engaging.
Gillian Diebold: What are the benefits of generative AI and Koalluh’s platform to students?
Varun Gulati: Many students are totally checked out of reading. In the United States, only 17 percent of kids read for fun. That’s alarmingly low, and this compounds the problem of declining literacy, which currently sits at 20-year lows. Koalluh makes reading way more engaging for kids by empowering them to create hyper-personalized stories and just-right text using the power of AI. On Koalluh, students can select their interests, such as characters and settings, and AI will generate a totally unique, illustrated story within seconds. This personalization not only sparks students’ imaginations and taps into their curiosity, but it also helps to build students’ desire to read. With generative AI, we can make reading deeply relevant and super engaging for each and every student.
Diebold: What are the benefits of Koalluh to educators, and how do you handle potential resistance?
Gulati: Teachers across the country are experiencing friction with the “science of reading” curriculum adoption—though it’s effective, much of it feels overly scripted or disengages kids. Koalluh helps educators by giving them a way to supplement their existing language arts curriculum with super engaging reading practice without having to compromise on research-backed practices.
In the beta product that we just released a few weeks ago, we tested with a few hundred students in the Atlanta, Georgia region and across Northern California. Teachers absolutely loved how engaged their students were. They mentioned that they loved that their kids could be creative and were given a lot of agency. And they appreciated that Koalluh was good for all types of learners—English as second language learners, struggling readers, advanced readers, etc. We’ll continue to expand the capacity for educators to use Koalluh for differentiation and targeted reading practice.
Diebold: How can tools like Koalluh help the learning process become more dynamic?
Gulati: Tools like Koalluh can help to make learning in K-12 classrooms more engaging, more personalized, more differentiated, and more accessible—a combination that has previously been exceptionally hard to do at scale. For the first time ever, we can meet every student exactly where they’re at with hyper-personalized content that meets the specific needs of each student, engages them, and is effective, comprehensible, and decodable.
Every educator wants to engage their students in their learning. Research backs up the value of engagement: we have evidence that improving student motivation and student engagement improves reading outcomes. Koalluh makes reading and writing more engaging and fun for students, which can not only foster a love of reading and writing but improve reading proficiency as well.
Diebold: What types of data do you need to build generative AI tools for the classroom?
Gulati: To build great generative AI tools for the classroom, you need two types of data. The first is good training data. In the case of Koalluh, this means quality stories and relevant stories that kids enjoy. The second is quality input for each individual generation. This means the inputs to the hyper-personalized stories. Koalluh gathers the interests of each and every student, as well as any parameters for the stories and text (phonemical patterns for the kid to practice, vocabulary, etc.), along with other data that our models generate.
When done effectively, data can be used to create personalized content that meets the specific needs of each student, as well as helps teachers make progress toward their learning objectives. Additionally, over time, the more data that is collected, the better the platform can become at providing personalized content.
Diebold: With the dawn of generative AI and personalized learning, will this be the end of collaboration in the classroom?
Gulati: Generative AI and personalized learning do not at all mean the end of collaboration in the classroom! There are many examples of generative AI tools that have allowed people to collaborate better—so why shouldn’t we bring that to students as well?
When used appropriately, generative AI tools like Koalluh can truly foster collaboration and creativity in the classroom. For example, Koalluh provides a social reading experience where students can not only create hyper-personalized stories but also share and read stories generated by their classmates. This immediately helps to build a sense of community and engagement in the classroom.