[LewksWongToMe] LayeredLearning

Home Page of LayeredLearning
Explanation to a 5 Year Old
Explanation to a PhD Candidate

Inspiration

During university, we've oftentimes found it difficult to understand various concepts due to no prior experience, leading to us falling behind during lectures. We believe there must be a faster way to digest information, and videos by WIRED showcasing topics explained in 5 different ways served as inspiration to develop LayeredLearning.

What it does

LayeredLearning is an OpenAI-powered educational platform that generates explanations and quizzes on virtually any topic. By providing context-appropriate explanations suitable for different learning levels, such as that of a secondary school student, it significantly speeds up the learning process, making even the most complex concepts easily digestible.

Additional subtopics are also provided for further reading, as well as a short quiz regarding the topic based on a preferred education level.

The link to the website can be found here.

How we built it

The backend was built using Flask web framework in Python, integrating the OpenAI language model for our core explanatory and quiz generation functions. We also used Firebase for our backend data storage, and implemented error handling and secure data management practices. For the front-end, we used Vue3 to ensure a responsive and interactive user interface.

Challenges we ran into

Initially, we wanted to integrate PDF and mp3 into our use case. However, it served to be difficult as the libraries tried weren't applicable to our use case, and developing an in-house model within 2.5 days wouldn't be accurate. Therefore, we decided to pivot into maximizing the value provided by our concept explainer feature.

Accomplishments that we're proud of

Given how difficult and potentially varied the responses by the OpenAI API can be, it was definitely reassuring to obtain specific results for our particular use case. The output helped summarize complex topics in the field of machine learning among other concepts, and is definitely something we believe can be expanded on beyond the scope of these 2.5 days.

What we learned

Through the development of LayeredLearning, we learned about the intricate process of integrating AI with web-based services, as well as managing user data securely. It taught us the importance of user-focused design and the role of natural language models in transforming education, as well as the importance of data security in our evolving world.

What's next for LayeredLearning

User Profiling: Creating a user profile and understanding the learning needs of the user, adjusting the model to their learning needs on top of explaining it in different levels of complexity.
E-learning course recommendation on platforms such as Udemy and Coursera. After the concept has been explained, should the learner want to understand about the topic(s) in intricate detail, LayeredLearning will recommend a few courses suitable for them.
PDF to text functionality so it can serve as a report summarizer
Speech to text functionality for summarizing key concepts highlighted in lectures, podcasts or videos

Additional Links: