Inspiration
We wanted to help community health workers capture a patient’s story clearly, especially in busy clinics where important details can get lost.
What it does
CareBridge organizes selected symptoms, onset, and the patient’s own words into a shareable handoff. It also highlights a small set of potential emergency warning signs and explains each prompt.
How we built it
We created a responsive browser demo using HTML, CSS, and JavaScript. Its transparent rules run locally, with no server or AI model.
Challenges we ran into
We needed to make the tool useful while avoiding the impression that it can diagnose or rule out danger. We kept prompts limited, explained their triggers, and made clinician review central.
Accomplishments that we're proud of
We built an offline-friendly prototype that preserves the patient’s own words, creates a structured handoff, and clearly shows how its prompts are triggered.
What we learned
Healthcare tools need clear explanations and human oversight. We also learned to describe CareBridge accurately: it demonstrates an explainable AI workflow, but does not currently use machine learning or generative AI.
What's next for carebridge
We’d co-design with patients and community health workers, then evaluate handoff quality and prompt accuracy using fictional cases before considering any real-world pilot.
