Artificial Intelligence for Community Health Workers (YEDI eligible)
David Levine, Professor
Business, Haas School
Applications for Fall 2024 are closed for this project.
Roughly 1.5 billion poor people receive healthcare from roughly 1.5 million community health workers (CHWs). CHWs are typically women with limited education and minimal training. ChatGPT, GPT4 and their peers should be able to provide high-quality support, especially if trained on the local clinical guidelines.
If you are selected for this research project, you have the OPTION to join the Youth Equity Discovery Initiative (YEDI) program. YEDI is a mentorship program that supplements your URAP research work with weekly workshops in both Fall and Spring semesters. YEDI creates a cohort of undergraduate scholars (usually around 20-30, many from underrepresented backgrounds) that are working on youth equity research projects, such as this one. The year-long program utilizes cascading mentorship, skill-building workshops, and professional development to provide supportive pathways into research careers, service, and leadership addressing adolescent inequities and well-being. Scholars greatly benefit from (and enjoy!) the extra training and support, as well as the professional development opportunities (including faculty guest speakers, graduate student and practitioner panels). The program culminates in a research presentation at the end of the program. The 2024-2025 YEDI cohort will meet on Tuesdays from 4:30 - 6:00pm beginning September 24th. To take part in the YEDI program, you MUST be able to attend the workshops at this time. As a YEDI scholar, you will sign up for a PH198 class and receive 1 unit. This is in addition to the units you receive for your URAP research project. For any questions about YEDI, please contact Marieka Schotland at mschotland@berkeley.edu.
Role: 1. Some programmers will create a searchable database of clinical guidelines and write an app that interfaces with large language models to search for the relevant clinical guidelines and use them to step a CHW through a patient's case.
2. Some interns with medical interests can work with health care professionals here and in India or Kenya to test quality of replies, create a test suite of artificial patients, and interview on usability.
3. Some interns with good connections to rural India, Bangladesh, Kenya or Uganda and appropriate language skills will interview community health workers (called CHWs, CHPs, ASHAs) and other healthcare professionals about the project.
Qualifications: An interest in health care (e.g., a pre-med), programming, an interest in using artificial intelligence to help the global poor.
Careful attention to detail and work with little supervision.
Hours: 6-8 hrs
Social Sciences Arts & Humanities