Kernel Quest

Inspiration

Our inspiration for Kernel Quest came from recognizing the different communities around us and how each student comes from different educational backgrounds, experiences, and levels of access to learning resources. Not every child has the same opportunities to develop their reading and comprehension skills, and these differences can contribute to a growing literacy gap.

We wanted to create a tool that helps bridge this gap by making reading more accessible, engaging, and enjoyable for elementary school students, particularly those in underserved communities who may not have access to individualized reading support.

We also recognized that children are multi-modal. Some learn best by reading, others by listening, speaking, or engaging with visuals. We wanted to create a space where students could practice their literacy skills in ways that feel comfortable to them, build confidence in their abilities, and develop the critical thinking skills they need to understand the world around them.

What It Does

Kernel Quest is an interactive literacy app for students in grades 3–5, with tools for both students and teachers.

Students can:

Read illustrated stories written for different grade levels.
Practice comprehension, vocabulary, and using evidence from a story.
Explore supports and activities such as read-aloud, word help, writing, drawing, and Word Rescue.
Earn points, stars, and badges as Poppy grows.
See their progress and revisit reading activities. Teachers can manage classes and student profiles, review class and individual progress, and prepare reports for families. Reports open as drafts in the teacher’s email app, so teachers can review and send them themselves.

How We Built It

Kernel Quest has a React frontend built with Vite and a small Node.js server for its AI features. The server coordinates agents for story support, tutoring, fluency, and teacher insights using Anthropic’s Claude API. The app also has built-in fallback behavior, so it can run in a demo mode without an API key.

We organized the project around two connected experiences: a student space made for reading and exploration, and a teacher space for understanding how students are progressing. Student and class data is stored in the browser for the demo.

Challenges We Faced

One of our biggest challenges was finding the right balance between gamification and meaningful learning. We wanted Kernel Quest to feel fun and rewarding without distracting students from reading comprehension itself. Every interactive element needed to support the learning experience rather than overwhelm it.

Another challenge was recognizing the wide variety of learners we wanted to support. Students come from different educational backgrounds, communities, and demographics, each with their own experiences, strengths, and learning needs. We wanted to create a tool that could meet students where they are rather than assume everyone learns at the same pace or in the same way. This challenged us to think about how to make reading practice approachable, flexible, and encouraging for students with different levels of literacy and access to educational resources.

Integrating AI into a practical educational workflow was another important consideration. We wanted Claude to support teachers and personalize the experience where appropriate, rather than make the application feel like just another chatbot.

What We Learned

Building Kernel Quest taught us to approach technology from the perspective of the people who will actually use it. An educational tool needs more than exciting features—it must help students learn, provide teachers with meaningful insights, and fit the needs of the communities it serves.

Our work with Science Extravaganza reinforced that students often have the creativity and problem-solving ability to succeed, but may need better support with the foundational skills that help them access opportunities. Literacy support can strengthen learning across subjects, including engineering, science, and everyday problem-solving.

We also learned that AI can support existing educational systems without replacing the people who make them effective. Teachers and families remain essential to a student’s growth, while technology can help them collaborate and provide support more effectively.

What's Next

We hope to continue expanding Kernel Quest to support an even wider range of learners and communities. Future improvements may include features for students with dyslexia, multilingual reading and language-learning support, and more accessible ways for students with motor disabilities to interact with the platform. We also hope to test Kernel Quest with educators and students in real classroom and after-school settings. Their feedback would help us improve the stories, activities, accessibility features, and teacher tools so the platform reflects the needs of the communities it is designed to serve. By continuing to build around the diverse needs and experiences of students, we hope to make literacy education more inclusive and help every child build the confidence to read, communicate, solve problems, and express themselves in their own way.