Inspiration

As students and TA's we have seen both how AI can be an amazing tool to study as well as one that can hinder learning and even lead to cheating. The other issue with using LLM's to study is that they can lose context, especially in long chats, which can lead to frustration as the chat quality deteriorates. We wanted to create an application that not only enhances learning but encourages it and always has the context you want for a given class. Another aspect we looked at was accessibility as some individuals may not be able to take notes, or have other needs that might make classroom learning harder for them. This application would allow them to interact with class materials in a way that would be similar to talking with the professor themselves.

What it does

Users can upload textbooks and other sundry class materials to a given class, during which all the material is vectorized. User inputs are then also vectorized, allowing for a similarity search among materials. This will then provide context to LLM in order to respond with the retrieved context. This means that the LLM will always answer relative to the materials from the class which can be important as things taught can vary from education level to education level as well as from professor to professor.

How we built it

We built it by leveraging our own individual strengths an skillsets as well as ensuring that a large part of the application was built on technologies we already had experience with to ensure a solid foundation upon which we could integrate the new technologies we wanted to use.

Challenges we ran into

Learning new technologies in a 24 hour timeframe as well as some inevitable technology glitches cause some consternation and headaches for us.

Accomplishments that we're proud of

We are super proud of creating something that is not only cool but creating some that we would like to, and will, use ourselves.

What we learned

That programming is a team sport. We also relearned that paired programming is a valuable tool when facing a tough problem as well as when learning new technologies in a short timeframe.

What's next for All In Education (aiEducation) Study Buddy

While the application works we want to polish it up and enhance the functionality even further to be a more robust and complete application.

Built With

Share this project:

Updates