InterLect: Where Asynchronous Lectures Meets Engagement

Inspiration

Asynchronous online learning sucks. One of the big issues is static prerecorded lecture videos. You can't ask questions easily, and they are entirely unengaging.

What it does

We make it so that you can take a YouTube video lecture, and quickly have a transcript, segmented video chapters, open ended questions, and more! This all appears in a web app, where learners can watch a lecture video, ask the lecture video questions, and be engaged as the lecture video asks them questions! All of the analytics are made available to the instructor.

How we built it

Python Flask, JavaScript, HTML, CSS. OpenAI's Whisper and GPT-4.

Challenges we ran into

No major problems. There are always more technical problems to solve, but it works great!

Accomplishments that we're proud of

It works great! I am ready to integrate it into a learning platform.

What we learned

I learned lots of interesting things about generative AI, it's strengths, and its limitations.

What's next for InterLect: Where Asynchronous Lectures Meets Engagement

We want to target synchronous online learning as well to bring the same great experience to live online lectures. This would include live transcripts, the ability to jump back in the lecture, and the ability to ask questions with immediate responses, without waiting for the instructor to be available during the lecture to address your question.