OpenTeacher - Hackathon Responses
Inspiration
I kept thinking about how unfair it is that quality education depends so much on where you're born or how much money your family has. Like, there are brilliant kids everywhere who just don't have access to good teachers or learning resources. So I thought - what if everyone could have their own AI teacher that actually gets them and helps them learn at their own pace?
What it does
OpenTeacher creates personalized AI teaching buddies for anyone who wants to learn something new. The AI figures out how you learn best, what you're interested in, and creates lessons just for you. It can teach in different languages, explains things in ways that make sense to you, and actually remembers what you've learned before. Think of it like having a really patient, smart friend who's always ready to help you understand stuff.
How we built it
We used bolt.new to rapidly prototype the frontend - honestly saved us so much time during the hackathon! For the AI backend, we integrated Google's Gemini SDK which gave us really solid natural language capabilities right out of the box. Built the whole thing with React and JavaScript, and Gemini's API made it super easy to create conversational teaching interactions. We focused on making it web-based so it works on any device with a browser, and kept everything open-source so other developers can jump in and improve it.
Challenges we ran into
Getting the AI to actually teach rather than just dump information was harder than expected. Also making sure it doesn't accidentally say something culturally insensitive or wrong took a lot of testing. Oh, and making it work well on older phones with slow internet was a pain - had to compress everything way down.
Accomplishments that we're proud of
We actually got it working! Built a prototype that can teach basic math and reading in English and Spanish. A few friends tested it with their kids and they were genuinely engaged, which was awesome. Also managed to get the whole thing running on a phone that's like 3 years old.
What we learned
Teaching is way more complex than it looks - there's so much psychology and relationship-building involved. Also learned that making AI accessible isn't just about the tech, it's about understanding what people actually need and how they live. Open source development is amazing when it works, but coordinating with volunteers across time zones is tricky.
What's next for OpenTeacher
If we keep working on this, we'd love to add more subjects, support way more languages, and maybe partner with some schools or NGOs to actually test it in real classrooms. The dream is still to get this to kids who need it most, but we know that's a marathon, not a sprint. For now, we're just excited we built something that might actually help people learn.
Log in or sign up for Devpost to join the conversation.