Inspiration
In the midst of midterm season, we, a group of computer science students, found ourselves drowning in a sea of hours-long lecture videos, tirelessly searching for specific topics to review. Tired and burnt out, we wanted to come up with a way to study efficiently.
What it does
CourseLook is a cutting-edge educational app designed to make students' lives easier when they need to find specific information in video lectures. Whether you're a computer science student with numerous lecture videos to go through the day before an exam or simply an eager learner, CourseLook has got you covered. This app uses advanced video analysis and search technology to help you extract knowledge from your video content efficiently.
How we built it
We used React and ChakraUI to build the frontend of CourseLook. On the backend we used the open source all-MiniLM-L12-v2 sentence transformer and cache the transcripts as well as the generated embeddings. This was done using a FastAPI backend
Challenges we ran into
The time constraints were tough, especially since we discussed heavily on what idea we wanted to build.
Accomplishments that we're proud of
We're proud to have built a fully functional web app in such little time, especially an app that could help us and other students in their studies. As students ourselves, we understand the difficulty of managing academic workloads, so we're hopeful that this will help others.
What we learned
We learned developing a fully fleshed full-stack application from scratch, including writing models to analyze video content, making the UI as simple and accessible as possible, and working together as a team. We also learned how to use open source embedding models and variants of vector search databases.
What's next for CourseLook
Sleep. And then deploy the website as a chrome extension so it's easily useable across youtube.
Log in or sign up for Devpost to join the conversation.