BrainBinge

• 

Inspiration

BrainBinge was inspired by my learning style. I’ve always learned best through analogies — connecting new concepts to things I already understand. When I relate a difficult idea to something familiar, it becomes much easier to absorb.

Over time, this turned into a habit of connecting school topics to my favorite shows and media. Even though studying and entertainment seem like they don’t belong together, I wanted to build something that connected the two. BrainBinge became that middle ground: a fun, engaging way to learn new subjects through familiar stories.

What It Does

BrainBinge lets users take any subject or topic they want to learn and combine it with their favorite show, anime, or media.

Users can:

• Enter a subject and a specific topic

• Choose their favorite show

• Select the type of questions they want (MCQ, free response, or mixed)

• Generate either a full explanation + quiz or a quiz-only mode

• Answer questions, get instant feedback, and earn XP

• Level up as they practice

• Everything is themed around the show they choose.

How I Built It

I built BrainBinge using Replit’s AI Agent as the core engine for generating explanations, analogies, and quiz questions. I chose to build it as a Flask-based web app because it was the simplest way to get a working product quickly.

I also used:

• GitBook for documentation

• GitHub for version control

• Canva for visuals and branding

The main workflow involved creating backend endpoints using Flask, sending structured prompts to the AI model, and building the frontend with HTML, CSS, and JavaScript to display interactive questions and XP progression.

Challenges I Ran Into

One major challenge was dealing with inconsistencies from the AI agent. Sometimes:

• Explanations wouldn’t show up

• The questions came out incorrectly

• The agent claimed something worked when it actually didn’t

Another challenge was building the free-response grading system. I wanted open-ended questions to feel fair, but I initially had no idea how to compare user answers to the model answer without being too strict or too lenient. Implementing similarity scoring and an XP system helped make grading more flexible and user-friendly.

Accomplishments I'm Proud Of

I’m really proud of how well the entire project came together. The UI and style feel clean and inviting, and the AI-generated questions stay surprisingly close to show “canon”. The analogies feel natural, not forced, and math and other subjects genuinely become more fun through show-based explanations.

Seeing random math questions rewritten using my favorite anime characters is something I didn’t think would work this well.

What I Learned

This challenge taught me how to use several new tools and technologies:

• I gained experience with Replit, especially handling AI agents

• I learned how to push code to GitHub and fix issues using the shell

• I learned documentation through GitBook, which was confusing at first, but became enjoyable once I understood it

I also improved at debugging, problem-solving, and searching for solutions independently.

Overall, I learned how to build a complete, functioning app from scratch with tools I had never used before.

What's Next for BrainBinge

There are several features I want to add next:

• A more advanced leveling system

• User profiles

• History tracking so users can see their progress over time

• A focus mode (like a Pomodoro timer) to help with study sessions

• More XP rewards, achievements, and long-term goals

The ultimate goal is to make BrainBinge not just a study tool, but a place where learning feels as exciting as watching a new episode of your favorite show.

Built With

• canva
• flask
• gitbook
• github
• javascript
• python
• replit