Skip to content

Welcome to VHMBot

Contents

This website contains documents regarding the policies, guidelines, and procedures for the Vanderbilt Section of Hospital Medicine. These can documents can be manually navigated via the contents bars to the right and left or searched via the search bar at the top. A generative AI powered chatbot accessed to the lower right is trained off of this knowledge base.

Purpose

This website is a branch of the XLR Project. The primary goal of XLR project is to provide more seamless and efficient access to institutional knowledge bases through generative AI integration. These knowledge bases have the dual benefit of trust: being authored by institution members, and location specific information that would not be available via other means. Please visit the website for more information.

Notes:

This website is maintained by Matthew Sweeney, MD and Chase Webber, DO through the XLR Project. The XLR Project is funded by the Innovation Catalyst Fund. The conversion from source files to the .md files displayed on this website is imperfect. Care has been taken to ensure the most accurate translation possible but there may still be errors both in content and formatting. The documents contained on this website are incomplete and in some cases out of date. They are being updated regularly.

And finally: WARNING

This website does not represent the official repository of VHM documentation. This website and its chatbot functionality are clinical decision making support tools and are in beta. The chatbot has been designed specifically to minimize hallucination and provide answers that are a direct match to the source materials within this website. Despite this there is always the possibility that an answer is incorrect or pulled from an outside knowledge source. Do not use answers provided by the chatbot as your sole guide in making important clinical decisions. Again: if in doubt, fact-check answers against the primary document source (Sharepoint or Teams) or another reliable source of clinical information.