AutoLearnr

Achieve mastery with Autolearnr: scalable personalized education for all.

Comment

Inspiration

Our inspiration for AutoLearnr came from a simple yet powerful idea: to democratize access to education and make personalized learning accessible to everyone. We were driven by the belief that traditional education systems often fall short in catering to individual learning styles and paces. Of course, how could they not? With over 1.5 billion students learning at various levels around the world, the resources necessary for customizing and catering to their individual needs would have been impossible. Yet, with the newfound capabilities of large language models, we were inspired to create a solution that would harness the capabilities of AI to provide a scalable and personalized educational experience. Just as everybody is different, so are the ways that they learn, and we believe this is the first time that anybody can learn anything, however they want.

What it does

AutoLearnr is an AI-powered learning platform for everybody. The goal: to simulate a teacher who has their attention solely focused on the student and how they learn. AutoLearnr allows you to submit class notes, textbook pages, or other documents, and generates a video that approaches the content in the best manner. While our AI tutor auditorily explains the most important points, students are able to view a slideshow with the most important information. Furthermore, students are able to consistently access assessments that measure their capability, as well as access an AI tutor that can answer their questions.

How we built it

We primarily leveraged the GPT 3.5 OpenAI API as a method of synthesizing and summarizing large amounts of text, as well as generating the script for the lesson. This, in combination with optical character recognition for pdf reading as well as utilizing text-to-speech from google cloud and a generated slide deck, form what we call our educational videos for any subject.

Challenges we ran into

Because a lot of the development of various features was done separately, integrating the different parts into one workable application was quite difficult. Another struggle we had was in the creation of the video, finding ways to sync the sound narration with the slides that would appear.

Accomplishments that we're proud of

Successfully creating a skeleton platform that simulates what an individualized learning experience would be like.

What we learned

Large language models have enormous abilities in synthesizing and summarizing information from large amounts of text and, in combination with other technologies, have the potential for creating truly individualized education.

What's next for AutoLearnr

Our aim is to create an authentically individualized comprehensive educational experience. Next, we hope to generate graphics, diagrams, and other visual mediums that can supplement the notes and audio that are already generated. Although we already measure student ability, we hope to use the data and information resulting from student performance in informing the content and lessons that the students receive, targeting their weakest subjects as areas of growth.

Built With

  • next.js
  • ocr
  • openai
  • text-to-speech
Share this project:

Updates