Inspiration
Learning a new language is hard. Watching YouTube is fun. We wanted to create a tool that makes language learning fun, practical, and personalized. The idea of using YouTube—a platform rich with diverse and engaging content—sparked our vision for MumboJumbo. By combining auto-generated curated questions, we aim to empower learners to dive into real-world language usage at their own pace and proficiency level.
What it does
MumboJumbo transforms any YouTube video into an educational experience for language learners. Users can select a video, choose a language they want to learn, specify their proficiency level, and how many questions they want to be asked. The tool then generates tailored questions to test comprehension and vocabulary based on the video content. Whether you’re a beginner or advanced learner, MumboJumbo adapts to your needs and offers an engaging way to improve language skills while enjoying videos.
How we built it
We built MumboJumbo using: Frontend: We utilized React for our UI, providing, video playback, and our questions. Backend: Node.js with Express handles API requests and integrates with external services. Language Processing: AWS Bedrock generates curated questions based on video transcripts and user preferences. YouTube API: Retrieves video metadata and captions to extract content for language learning.
Challenges we ran into
A: First timer with the YouTube API. RIP 4 Hours of no progress. K: Figuring out how to get the correct questions displayed at the right time was interesting. B: Figuring out front-end styling and animations S: Prompt-engineering. is hard
Accomplishments that we're proud of
We have a working application and proud of that. Also learned some new technology.
What we learned
Some cool new technology.
What's next for MumboJumbo
We can expand our suite of question types, as well as cater to other media content (such as live videos).
Log in or sign up for Devpost to join the conversation.