CORE Medical Communication Console

Inspiration

Healthcare communication often begins with uncertainty. Patients and families may receive an explanation but still struggle to understand what has been decided, what remains unclear, and what they should ask next. Healthcare teams, meanwhile, need information that is organized, reviewable, and careful not to replace clinical judgment.

CORE Medical Communication Console was created to explore a safer bridge between these two sides. It grew from practical experience in hospital public relations and healthcare communication in Japan.

CORE is not an AI doctor. It is a communication preparation and handoff tool.

What it does

CORE is an interactive prototype that turns patient and family concerns into structured information for human review and follow-up.

For example, a patient’s family may leave a medical explanation unsure about what has been decided and what they should ask next. CORE organizes that uncertainty into questions to confirm, a summary of shared understanding, and items to hand off to the care team.

The console guides users through a fixed scenario and demonstrates how unclear concerns can become a more organized starting point for the next human conversation.

CORE does not diagnose, prescribe, or make medical decisions. The current prototype uses no real patient data. Its purpose is to demonstrate communication design, information structure, and a human-in-the-loop safety model.

How we built it

The prototype was developed through an iterative process of dialogue, specification, implementation, review, and deployment using Codex and GPT-5.6.

Existing healthcare communication knowledge and materials were reorganized into an interactive console. Codex supported code inspection, implementation, interface refinement, consistency checks, and deployment preparation.

Human review remained responsible for the project’s purpose, wording, safety boundaries, and final decisions. This division of roles was intentional: AI helped transform domain experience into a working prototype, while the human creator retained responsibility for meaning and safety.

The application was built with React, TypeScript, and Vite, and deployed on Vercel.

Challenges we ran into

The main challenge was balancing clarity with safety. The interface needed to help users organize difficult concerns without appearing to diagnose, prescribe, or replace a healthcare professional.

We also needed to translate healthcare communication practice into a small and understandable prototype. The experience had to make sense not only to healthcare professionals, but also to people unfamiliar with medical communication.

Another challenge was making the boundary between AI assistance and human responsibility visible throughout the interface, rather than relying only on a disclaimer.

Accomplishments that we're proud of

We created a working public prototype that demonstrates an end-to-end communication flow rather than presenting only a concept or slide deck.

The project shows a clear transformation:

unorganized concern → questions to confirm → shared understanding → human handoff

The prototype makes its limitations visible, uses no real patient data, and keeps human review at the center of the experience.

We are also proud that this project demonstrates how a healthcare communication professional without a traditional software engineering background can collaborate with Codex to turn domain experience into a testable digital product.

What we learned

We learned that AI can provide value before a clinical conversation—not by making medical judgments, but by helping people organize concerns, identify unanswered questions, and prepare information for human review.

We also learned that safety is not only a disclaimer. It must be reflected in the interface, language, workflow, and handoff design.

Finally, we learned that Codex can function as more than a coding assistant. It can help connect domain knowledge, product structure, implementation, and iterative review while leaving important decisions with the human creator.

What's next for CORE Medical Communication Console

The next step is to test the communication flow with healthcare professionals and refine the terminology, accessibility, and handoff process based on their feedback.

Future development may include additional communication scenarios, multilingual support, stronger review records, and carefully governed integrations.

Any expansion will continue to preserve three principles:

• CORE supports communication rather than making clinical decisions.
• Human professionals retain responsibility for review and care.
• Real patient data will not be used during prototype testing without appropriate governance and safeguards.

Built With

• codex
• github
• gpt-5.6
• react
• typescript
• vercel
• vite