PeerConnect AI

A lite weight tutoring tool for students in Math, Physics, Chemistry, and History.

Comment

Inspiration

Online tutoring platforms often focus on scheduling sessions, but not on making the learning itself more interactive or adaptive. Many students struggle to stay engaged or get feedback in real time, especially when studying alone. PeerConnect AI was inspired by the idea of giving every learner their own responsive, conversational tutor — one that can explain, quiz, and adapt its tone like a real teacher.

What it does

PeerConnect AI is an AI-powered tutoring assistant that adapts to a student’s learning mode. It features:

• Explain Mode – breaks down complex topics step-by-step in plain language.

• Quiz Mode – tests understanding through quick, interactive questions.

Together, these create a seamless tutoring experience — available anytime, from any browser.

How I built it

PeerConnect AI was built as a React + Vite + Tailwind web app connected to the NeuralSeek API.

• The frontend uses Tailwind’s prose typography and react-markdown + rehype-katex to render rich explanations, formatted text, and math equations.

• A simple mode and persona system controls how NeuralSeek responds — switching between tutoring and quizzing modes dynamically.

• React Hooks manage state for chat messages, scroll behavior, and context memory.

The result is a lightweight, responsive, and highly customizable tutoring interface.

Challenges I ran into

• Getting the API key to load properly without exposing it required reworking environment variables.

• Interfacing with a NeuralSeek Agent through Vite

• Rendering LaTeX equations inside chat bubbles cleanly took trial and error with Markdown and KaTeX setups.

• Balancing explanation vs. quiz tone required fine-tuning the prompt logic to make AI responses sound human and consistent.

Accomplishments that I'm proud of

• Built a fully functional AI tutor with two learning modes — entirely solo — in under 36 hours.

• Achieved smooth math and Markdown rendering for professional-looking AI explanations.

• Created a clean, minimalist UI that feels natural to use on both desktop and mobile.

What we learned

• How to integrate NeuralSeek’s API with Vite and Node.

• Creating AI Agents and Agent Flow

• The importance of prompt design and persona-based instructions for consistent AI behavior.

• Practical use of React Markdown + KaTeX for displaying rich educational content dynamically.

What's next for PeerConnect AI

• Adding progress tracking to visualize student improvement.

• Implementing voice feedback and speech input for hands-free learning.

• Expanding into collaborative peer study rooms, where multiple users can learn with a shared AI tutor.

• Instructor dashboard to monitor student progress and understanding.

Built With

Share this project:

Updates