We will show a chatbot that predicts diagnoses from chest X-rays using MONAI Toolkit’s TransCheX model and Hopsworks.

The chatbot also summarizes medical reports and presents them in an easily understandable format. We will go under the hood of our chatbot to show you how the chatbot provides accurate and reliable predictions quickly at any scale and with zero downtime. For example, we will explore the functionality of Hopsworks in processing and serving both structured data (patient records) as well as unstructured data (XRay images) in real-time. The chatbot uses function calling capabilities to retrieve patient information and make predictions based on the conversation.

Finally, we will also discuss how to use large language models (LLMs) for summarizing and presenting medical reports.