Inspiration
As a student myself, I understand how daunting it can be to grasp new concepts presented in voluminous and dense textbooks. Leveraging the impressive capabilities of large language models, I envisioned an assistant that simplifies study materials, making them more accessible and comprehensible.
What it does
Study Buddie, as the name suggests, is a companion designed to aid in studying. Its key features include:
- Chat Interface: Enables interaction with uploaded PDF study materials. Self-Evaluation: Offers objective and theoretical questions based on the material for self-assessment.
- Automatic Restoration: Restores previous conversations for seamless continuation of learning.
- Flashcards: Assists in retaining and recalling key concepts.
How we built it
We developed Study Buddy using the Python programming language and a combination of tools. Langchain was employed to enhance the accuracy of LLM responses. The backend authentication was powered by Flask, with MongoDB utilized for storing credentials and conversation history.
Challenges we ran into
One major challenge was the lack of capable image models to achieve desired visualizations for the app. For instance, a feature where the app generates a canvas with summarized key points proved unattainable. We hope for advancements in image models or eventual training of models to handle such tasks. Additionally, the limitations of Streamlit for frontend development posed constraints in extending and implementing certain features, given its Python-centric nature.
Accomplishments that we're proud of
We take pride in successfully implementing the self-test feature, which presented complexities in combining self-testing with feedback from language models. Furthermore, the automatic restoration of conversations underscores the app's role as an ever-present study companion, ready to assist at any time.
What we learned
Our experience highlighted the immense potential of AI as a tool to significantly enhance various tasks, including reading and studying. Text models offer a promising avenue to make studying more engaging and concepts more digestible. However, we also acknowledge the ongoing journey toward achieving intelligent capabilities, particularly in areas such as image recognition.
What's next for studyBud
In the future, we aim to further enhance Study Buddy by exploring advancements in image models and frontend development to provide richer visualizations and an improved user experience (which has been started). Additionally, we seek to refine and expand its features to cater to a wider range of study materials and learning preferences.
Log in or sign up for Devpost to join the conversation.