EquiLearn

Inspiration

Some of my friends (and oghers I observed) find it difficult to find success in finding internships networking scholarships or even trying go better. their grades due to lack of resources / underfunding. This is known as the "hidden curroculum" as students with less resources often don't realize or know how to actively search for opportunities that can enhance their quality of life and bring them more success.

What it does

This mini RAG system consists of an AI assistant that answers any user questions regarding internships, networking, any informstiom on how to do well/better in classes or scholarship informarion/applications

How we built it

We used a vector knowledge base that consists of all reliable sourcss used, a local LLM model hub called Ollama to run the assistant, streamlit for the local web interface, and python for provramming the assistant and interface chat logic.

Challenges we ran into

We had issues with the model taking too long to respond to user's questions, as well as finding a way to make the model avoid mixing different topics of information (ie. scholarships and internships).

Accomplishments that we're proud of

Applying my knowledge of topics in AI and software development to create a real-time project that will help any underprivileged student (ie. with a lack of resources).

What we learned

This was a first hackathon for some of us, so we learned to divide roles and ultilize github features to the fullest extent for efficiency of making edits and transferring files.

What's next for EquiLearn

Continue pulling more reliable information to make my model even more accurate with responses. Deploy on the internet for thousands of necessary users to help them navigate the hidden curriculum and have it make a positive mark on the community.