HealthAdvisor: GenAI-based personalized wellbeing app

GOAL

Personalized AI wellbeing advisor for behavioral change.

PROBLEM

Mental health crisis among the general population.

• For instance, nearly 50% of post-graduate students report some signs of depression or anxiety.
• Personalized healthcare advice is costly as it requires access to human experts.

SOLUTION

Personalized and actionable plans based on individual context.

• Augmented via physiological states/behavior, personal profile and health goals.
• Leveraging Generative AI.
• Conversational interface for follow-up questions.

Achieve wellness goals.

• Improve sleep, healthy weight range, improve motivation and nutrition intake, physical activity, and social support.

Text-only large language models usually struggle with complete context and an understanding of common sense knowledge.

• HealthAdvisor augments context by physiological and behavioral data from wearable sensors, self-reported health goals, and medical history.

HYPOTHESIS

GenAI-based personalized plans can outperform existing one-size-fits all approaches.

RESULTS

We avoided bias by applying a different model for Quality/Safety checks, which helped us to improve the outcomes in scenarios with poor intervention options.

We achieved more personalized interventions by utilizing GenAI and augmenting context, compared to the traditional one-size-fits-all approaches.

Team Members

• Anshul Madan (Lead) (madananshul@gmail.com)
• Eduard Pogorskiy (epogorskiy@acm.org)
• Shifali Sharma (shar2@umbc.edu)
• Sonish Sivarajkumar (sonish.sivarajkumar@pitt.edu)

FEEDBACK

Seeking feedback on HealthAdvisor project, happy to hear feedback over email or even have a short 30 minute call (e.g. how useful you think this product could be, other adjacent areas we could look at, or ideas on datasets we could use)

Mail: Anshul Madan madananshul@gmail.com

INTERVENTION EXAMPLES