MedAgent is an AI web app that helps medics to check data on clinic histories and databases, reducing the time needed for consulting. MedAgent is an AI web application built on the Streamlit framework, it uses langchain libraries for text extraction, division and chunk creation. After that is uses the Faiss algorithm to perform a semantic search based on the user’s (medic) input, so it finds the most relevant text chunks for the question. Finally it uses OpenAI api to generate a response based on the data found. It can be fed with a pdf clinic history or a db as a source. It is possible with more time to integrate a voice assisted querying and an automatic clinic report generation.
Category tags:"Great job applying AI to streamline healthcare information access - with some refining this could become a very useful tool for improving clinical efficiency."
Olesia Zinchenko
Product Marketing Manager