LinguaLoop

Inspiration

Language immersion works best when it fits naturally into everyday life. LinguaLoop starts with the short videos people already want to watch and turns unfamiliar words into small, in-context learning moments.

What it does

LinguaLoop is a mobile-first web prototype for learning from short videos. Users scroll through Spanish clips with synchronized captions, turn on English translations when needed, and tap a word to see its meaning in that sentence, hear its pronunciation, and save it without leaving the video.

Saved words become contextual flashcards: learners recall the meaning, then replay the sentence where they heard it. The prototype also includes a video tutor, a full transcript with sentence replay, local text scanning for text visible in a video frame, learning preferences, and progress tracking.

How we built it

We built LinguaLoop with Next.js, React, TypeScript, Tailwind CSS, and Lucide. Tesseract.js runs in a local browser worker for text scanning. The demo catalog contains three curated Spanish TikTok clips with prepared captions, definitions, source links, and excerpt timings. Progress and saved vocabulary are stored locally in the browser. The tutor uses built-in practice responses by default, with an optional AI provider for live answers.

Challenges we ran into

A key challenge was making vocabulary practice feel native to watching rather than a separate task. We designed the experience around context: every saved word keeps its original sentence, and learners can replay that moment while reviewing. We also kept clear prototype boundaries: this is not a live TikTok feed or universal language catalog, and local progress does not yet sync between devices.

Accomplishments that we're proud of

We created a complete learning loop around short-form video: watch, understand a word in context, save it, and practice it later in the same context. The prototype brings together video playback, timed captions, vocabulary, flashcards, tutoring, text scanning, preferences, and progress in one responsive experience.

What we learned

We learned that language practice can be more approachable when it builds on media learners already choose to watch. Keeping meaning, pronunciation, and review connected to the original conversation makes each interaction more relevant than an isolated word list.

What's next for LinguaLoop

Future directions include expanding the language and video catalog, enabling richer video importing, and adding cross-device progress syncing while preserving the context-first learning loop.