StudyBuddy.ai

Inspiration

When using standard AI tools like ChatGPT for studying, the responses can sometimes be inaccurate, misaligned with course materials, or even too advanced for the student’s level. We wanted to create an AI-powered study companion that truly follows the course syllabus and textbook ensuring that students receive reliable, level-appropriate help.

What it does

StudyBuddy.ai allows users to upload their course outlines and textbooks. The system then tailors its responses to match the specific course content, providing accurate, relevant, and course-aligned explanations. This ensures that students learn exactly what’s required for their class no more, no less.

How we built it

We integrated OpenAI’s models to analyze course guidelines and textbooks, combining this with a PostgreSQL database to store textbook data and chat history. This enables the AI to “remember” past conversations and build continuity in tutoring sessions.

Challenges we ran into

Uploading entire textbooks directly into the model consumed a large number of tokens, leading to inefficiencies and higher costs. Managing token usage while maintaining accuracy was one of our biggest technical challenges.

Accomplishments that we're proud of

We developed a smart retrieval algorithm that identifies the most relevant textbook chapters based on the user’s question. This dramatically reduces token consumption while improving both speed and precision, allowing StudyBuddy.ai to deliver focused, high-quality answers in real time.

What we learned

We gained hands-on experience integrating OpenAI with a frontend interface and learned how to use PostgreSQL effectively alongside AI models. We also deepened our understanding of optimizing token efficiency and building context-aware AI systems.