Bubbles AI

Main page
AI generation of problems based on difficulty and data structure
AI solution provided by AI with explanation, steps, and suggestions.
Landing page

Track

This track is for the HeadStarter AI agent track, where I made the AI agent into a Data Structure mentor for Hispanic people learning data structures. Also, this project is for the tracks best overall yay, most likely meme startup, lowkey actually kind of good.

Inspiration

The inspiration for Bubbles AI came from the desire to mentor data structures for Hispanic learners diving into computer science. I envisioned an AI-powered mentor that dynamically generates personalized problems to enhance learning.

What it does

Bubbles AI is an interactive platform that helps users practice data structures through personalized and AI-generated problems. The platform adapts to the user’s progress, offering challenges of varying complexity, hints, and step-by-step explanations. Users can track their improvement and receive instant feedback, creating a tailored and engaging learning experience.

How I built it

I utilized cutting-edge technologies to bring Bubbles AI to life:

LangChain: To create personalized and context-aware problem statements.
Groq: For generating diverse and accurate AI-driven problem sets.
Backend: Built to handle user authentication, progress tracking, and problem data storage.
Frontend: Designed to provide a seamless user interface for solving problems, viewing hints, and tracking progress.

The integration between LangChain and Llama3.2 ensures that the generated problems are both relevant and varied. The backend was optimized for scalability and speed to handle multiple users simultaneously.

Challenges we ran into

One major challenge was fine-tuning the AI to generate problems that are challenging yet solvable for users of different skill levels. Ensuring the questions were meaningful, varied, and matched the user’s progress required significant effort.

Another hurdle was integrating LangChain and Groq into the platform without compromising performance. I had to optimize the backend to ensure smooth and fast problem generation, even under heavy loads.

Finally, designing an intuitive and engaging user interface was a balancing act. I aimed to include all necessary features while keeping the interface clean and user-friendly.

Accomplishments that I am proud of

I am proud of creating a platform that:

Dynamically adapts to users’ skill levels, ensuring a personalized learning journey.
Seamlessly integrates advanced AI technologies to generate high-quality problems.
Provides an intuitive and engaging experience that motivates users to learn.

What we learned

Building Bubbles AI deepened our understanding of AI-powered educational tools and how to create adaptive learning experiences. I gained valuable insights into:

Implementing LangChain and Groq for problem generation.
Optimizing backend performance for scalability and speed.
Designing user-centric interfaces that simplify complex processes.

What's next for Bubbles AI

The journey doesn’t stop here. Upcoming enhancements include:

Adding support for more advanced topics, such as graph algorithms and dynamic programming.
Introducing competitive features, like leaderboards and timed challenges.
Expanding hints and explanations with AI-powered tutorials for deeper learning.
Building a mobile app to make the platform more accessible.