-
-
Home Page
-
List of courses for the student from canvas
-
List of courses the user added
-
List of files in that course
-
Study path generated for the specific file
-
Study path being used as a progress tracker
-
AI Tutor chat
-
AI Tutor chat giving customized feedback
-
AI Tutor adding references for responses out of memory
Inspiration
Navigating the multitude of files and notes scattered across different courses in Canvas often leads to study inefficiency and stress. We were inspired by the struggle to consolidate resources and visualize a clear path forward. This led us to create StudyBuddy — a tool designed to unify course materials and automatically generate a personalized, actionable study roadmap from the chaos of digital class files.
What it does
StudyBuddy acts as an intelligent assistant, connecting directly to the user’s learning management system (LMS). It enables users to fetch course listings and then select specific files (lectures, readings) to add to its dedicated knowledge memory.
The AI uses this contextual data to answer specific user queries, offer detailed summaries, and most importantly, generate a structured study roadmap tailored precisely to the user's learning materials.
How we built it
The application relies on key integrations to function. We utilized the Canvas API to securely access course data and retrieve specific files. These files are then sent to the Gemini API, which processes, indexes, and uses them as the grounding source for customized responses.
JavaScript handles the front-end user interface and the necessary logic for data management and user authentication.
Challenges we ran into
The primary challenge was managing the development lifecycle. We found ourselves constantly grappling with the difficulty of prioritizing new features (like advanced roadmap generation) against the need to maintain stability in existing systems (like secure file ingestion and API handling).
It was a delicate balance of trying to build "what's next" without inadvertently breaking "what's already there."
Accomplishments that we're proud of
We are most proud of the deep integration that transforms raw course files into actionable intelligence. Seeing the system successfully ingest dozens of documents and then output a highly specific, customized study roadmap — a true plan of action — was a major technical and functional achievement.
What we learned
The project provided a significant learning experience in complex API integration, specifically in securely handling user-specific educational data and marrying it with the contextual power of a large language model.
We also learned valuable project management lessons regarding scope control and the necessity of rigorous testing for dependencies between the Canvas data layer and the Gemini intelligence layer.
What's next for StudyBuddy
The future of StudyBuddy focuses on scaling and deeper integration. We plan to achieve multi-tenancy to support various user groups and institutions easily.
Feature-wise, we will be implementing automated flashcard generation from uploaded materials and greatly expanding our collaborative functionality to support group study sessions, allowing users to share resources, work on roadmaps together, and co-study in real-time.
Finally, we will continue enhancing the experience by gamifying the study process.
Built With
- claude
- python
- react
- sqlite
- supermemory
- typescript
Log in or sign up for Devpost to join the conversation.