The objective
An AI-powered chatbot designed to help pathologists quickly access relevant clinical research and evidence-based insights, delivering recommendations drawn from the latest medical publications.
Pathology general presentation
This project improves clinical research workflows for pathologists by using an AI-based system that retrieves and prioritizes relevant medical literature, replacing time-consuming manual review processes.
Technical approach
Named Entity Recognition (NER) and Retrieval-Augmented Generation (RAG) techniques were applied to extract clinical recommendations from top-ranked research articles and classify them by level of evidence: Strong, Moderate, and Low.
Results
The system generated concise recommendations while displaying source articles, personalized relevance scores, and complete metadata, providing pathologists with actionable insights and full traceability back to the original evidence.
