CareCoordinator

CareCoordinator is a private web application for families who need to keep complex medical care organized across people, locations, and communication channels. It began as a practical response to a close family member's serious diagnosis where an overwhelming amount of information kept arriving through messages, calls, documents, and specialist follow-ups without a single shared view of what had happened or what needed to happen next.

The application gives each family a separate workspace. Members can maintain a current summary, dated timeline, patient profile, referral tracker, reports library, activity log, decision tree, and assigned action items. New notes, dictated text, copied message or email threads, images, and uploaded reports can feed an AI-assisted processing pipeline that extracts structured updates. The app also provides workspace-grounded chat, natural-language document search, and research and surgery-note views that can be enabled per workspace.

CareCoordinator uses Next.js and React for the interface, Supabase for the application backend, and Anthropic for its AI features. Its configurable workspace model allows the same product to support different families, care situations, categories, and optional tools without hard-coding one patient’s details. Structured AI outputs help turn unstructured updates into predictable application data while keeping the original source material available for review.

I designed the product around the family coordination workflow and directed an AI-assisted implementation, including the multi-workspace product model, AI workflows, testing, and deployment process. The project is now running as a small private family pilot. It is an operational tool rather than a public medical service: AI-generated organization is intended to support—not replace—professional medical advice or the underlying clinical records.