Overview
This playbook provides a step-by-step guide for applying human-centered design (HCD) methods to create benchmarking datasets for large language models (LLMs) used in clinical decision support. It outlines how to involve frontline health workers in surfacing realistic medical scenarios, incorporate perspectives from underrepresented populations in data collection, and refine datasets through ongoing collaboration with local stakeholders.
Do you have feedback on this resource?
Thank you for your feedback as we strive to curate and publish resources to help social impact organizations succeed with data.
Related Resources
How to assess an organization’s data asset risk
In this guide, you will learn how to first identify your data assets, and then assess their associated risks.
How to improve data quality through validation and quality checks
This guide will explain data cleaning processes and their value in building the usefulness of your data.