Inspiration
We noticed that students often struggle with topics that are not covered in their course resources, while teachers may not know what students are struggling with. This inspired us to build CoursePulse, a platform where student questions can help improve the course itself.
What We Learned
We learned how to integrate Gemini AI meaningfully into an educational workflow, especially for course-grounded answers, knowledge-gap detection, and personalized explanations. We also learned the importance of clearly separating teacher-provided information from general AI knowledge.
How We Built It
Teachers upload course resources, which Gemini analyzes. Students can ask questions and receive either Course Grounded answers or Beyond Course explanations. Repeated questions reveal Resource Gaps, which teachers can review and turn into new AI-assisted learning resources. We built it with Next.js, Node.js, PostgreSQL/Neon, Prisma, and Gemini API, with deployment through Vercel.
Challenges
Our biggest challenge was time limitation . That's why the project was small enough .We focused on one core idea: creating a two-way learning loop between students, teachers, and AI rather than building another large LMS.
Built With
- express.js
- gemini-api
- google-ai-studio
- neon
- next.js
- node.js
- postgresql
- prisma
- react
- render
- tailwind-css
- typescript
- vercel
