Inspiration
Whilst digital learning is ubiquitous with more than 2.4 trillion hours being spent on it annually, most of it is plain video content served to people at different times creating the feeling of self pacing and control over the curriculum. But not everybody learns the same way through videos.
What it does
We want to provide people with a custom tailored learning platform, from which they can't just learn, but that also learns from them and adapt to their needs.
How we built it
This dream was able to come to live, through the use of state of the art LLMs and other AI-technologies that generate plans on how to teach subjects, certain chapters of said subjects and lastly also how to fill concrete lessons. These technologies are woven together using three different layers of interconnectivity, for this Python, Typescript NestJS, Next.js and React are employed.
Challenges we ran into
Our platform is built with safety and credibility in mind. This means we are building upon accessible technologies to ensure everyone can use it and education is democratized. Due to this constraint it took us a lot of time to fine-tuning our AI solutions and we had to switch models multiple times to achieve the best results
Accomplishments that we're proud of
We are proud of the amount of knowledge, new technologies and new friendships we were able ascertain.
What we learned
Beyond learning the hard way that sleep is not optional, we learned just how hard it is to ensure that AI outputs are consistent and truthful
What's next for LingoLearn
We hope to expand our current platform, which is currently amazing for language learning, but we would be thrilled to see people be able to learn more about disciplines people struggle with, such as history, coding and complex math.
Built With
- gcp
- nestjs
- nextjs
- palm
- python
Log in or sign up for Devpost to join the conversation.