innovation
A clearer next step, with a human boundary
A patient-navigation concept and prototype exploring clearer next steps in care.
- Role
- Concept and prototype developer
- Status
- Educational concept and prototype
Context and real problem
Navigating a care pathway can be difficult even when information is available. MedPal explores how an interface might make education, navigation, and escalation easier to understand without replacing clinical judgment.
Andres’s role
I developed the patient-navigation concept, mapped workflows, and worked on the prototype and validation plan. The prototype uses Next.js, Supabase, and OpenAI APIs.
Collaborators and setting
MedPal appears in my public Thought for a while learning repository and in my professional account of biomedical-innovation work. It is a learning and research project. This summary does not imply a hospital client, institutional endorsement, or a product already used in care.
Methods and tools
The concept considers patient and clinician workflows, education needs, navigation questions, and routes for escalation. Prototype work uses a web interface and language-model APIs. Workflow mapping and validation planning matter alongside the technical implementation.
Responsibility and safety
Privacy, health literacy, bias, access, and human oversight are design questions. A conversational interface must not make treatment decisions or be mistaken for clinical advice. No patient dialogue, health record, or fabricated screen is published here.
Outputs and outcomes
The output is a concept and prototype, documented as a learning project. Clinical validation and readiness for patient use remain unestablished.
What I learned
Navigation support needs a clear boundary between helping someone understand a next step and deciding that step for them. A useful prototype makes that distinction visible and gives people a route back to a qualified care team.
Current status and links
MedPal remains an educational concept and prototype. The repository defines that scope and states that its prototypes must not be used for real patient decisions.