Inspiration

Scientific research has shown that studying is only effective when it feels hard. But with generative AI like ChatGPT, Copilot, and Claude, it is easier than ever to let an AI do the work for you. And we get it, after staring at a math problem for eight hours, resisting the temptation is brutal. Students do not need an AI to do the work for them, they need guided learning, not shortcuts. That is why we built JammedAI, the study partner that keeps you thinking instead of checking out. It is designed for a variety of subjects such as math, physics, and chemistry. You snap a picture of your question, and in the background, JammedAI solves it but locks the answer away.

What it does

JammedAI is a studying and tutoring tool that gives smart hints, checks your work, and guides you step by step without revealing answers too soon. You upload a question, and JammedAI solves it in the background but keeps the solution locked. It supports layered hints, mistake detection, optional mini lessons, and ADHD friendly focus modes.

How we built it

We built JammedAI using React, TypeScript, Tailwind CSS, Vite, Express, and Google Generative AI. The frontend handles image uploads, hint loops, and UI customization. The backend processes problems, generates hints, checks work, and manages the solution unlocking logic.

Challenges we ran into

We had to design an AI workflow that does not give answers immediately, which is the opposite of how most models are used. Building the hint loop logic, detecting partial work, and ensuring hints were helpful without revealing too much was a major challenge. We also had to accommodate for ADHD and different focus styles. You can switch between calming, high‑contrast, or low‑stim light modes, and adjust pacing so the app works with your brain, not against it.

Accomplishments that we're proud of

With zero hackathon experience, we created a tool that genuinely supports learning instead of shortcutting it. We built a full hint loop system, integrated optional mini lessons, added ADHD friendly modes, and designed a workflow that keeps students engaged without overwhelming them.

What we learned

We learned how to control generative AI outputs more intentionally, how to design for different learning styles, and how to balance guidance with challenge. We also gained experience building a full stack app with real time AI interactions and user friendly UI patterns.

What's next for JammedAI

We plan to add progress tracking, spaced repetition review, support for more subjects, collaborative study modes. To make the videos more user friendly and entertaining, we plan to add a variety of cartoon characters to choose from. We also want to expand the lesson library and improve personalization so JammedAI adapts even more to each student’s learning style.