Inspiration
Literature review takes a long time due to the vast number of papers that are retrieved. So we wanted to save researchers some time...
What it does
Medical Research Explorer is a AI solution designed to streamline the exploration and analysis of medical research papers. It uses a combination of Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) to automatically summarize complex medical literature and answer user-generated questions. This tool is ideal for medical researchers, and students who need quick access to relevant information without manually sifting through extensive research papers.
How we built it
We built Medical Research Explorer by integrating a pre-trained LLM with RAG technology, allowing the system to retrieve relevant information from a vast database of medical literature. Our team developed a custom pipeline that scrapes research paper abstracts from medRxiv, extracts key insights from research paper abstracts and fine-tuned the LLM to ensure accurate and contextually relevant summaries. We also implemented a dynamic Q&A system that responds to user queries with concise, accurate answers based on the latest medical research.
Challenges we ran into
- Lack of experience with the databricks environment made it challenging to develop our application - particularly cost and resource constraints while calling LLMS.
Accomplishments that we're proud of
We're proud of our system, which consistently delivers concise and relevant summaries. Our ability to create a responsive Q&A system that can handle complex medical questions is a significant accomplishment.
What we learned
How to implement a RAG and provide context to LLMS. How to use databricks notebooks.
What's next for Medical Research Explorer
chunking documents so that it can handle longer research papers as opposed to just abstracts. Producing a more scalable infrastructure and enhancing UI and connecting and testing different LLMs.
Log in or sign up for Devpost to join the conversation.