Document search and content learning reimagined. The ultimate study tool, ScholarSearch provides intelligent document querying **and **conversational active recall, enabling you learn quickly and learn more.
12 lecture notes, each 30+ pages? Feeling unproductive with passive reading? Scholar Search offers a powerful search algorithm for your lecture notes, and an active learning approach for power learner!
Inspiration
Special thanks to Zilliz for providing the infrastructure! ScholarSearch leverages the powerul paradigm of Retrieval Augmented Generation (RAG) to provide a highly personalised learning experience.
What it does
ScholarSearch leverages Retrieval Augmented Generation to provide a highly personalised search engine. Query for a topic or general concept, and receive a detailed explanation that combines information across all of your notes.
Scholar Search also engages the user in active learning through using the Feynman method -- to practice a topic, engage in conversational active learning where the model prompts the user and evaluates their responses.
How we built it
The backend was built in Python (Flask) and uses a Milvus vector database. The frontend was developed using React.
Log in or sign up for Devpost to join the conversation.