Home screen with multilingual voice input for AI-powered nursing documentation generation.
Customize terminology, language, theme, and documentation behavior for different healthcare facilities.
Enter clinical findings by voice or text and select additional outputs such as SBAR and LAR email.
AI detects missing guideline-required fields before generating documentation to improve completeness.
Professional SOAP note automatically generated from clinical findings using facility guideline templates.
Automatically generates a family/LAR update email from the same assessment without duplicate documentation.
Built-in documentation quality check highlights completed items and missing required information.

Inspiration

As a registered nurse, I spend a significant amount of time completing clinical documentation. I wanted to create a tool that could reduce documentation burden and allow nurses to focus more time on patient care.

I created Nurse Buddy to transform natural, multilingual voice input into clear and professional nursing documentation.

What it does

Nurse Buddy is an AI-powered nursing documentation assistant that converts spoken or typed clinical information into structured documentation, including:

SOAP notes
SBAR reports
Progress notes
Family and LAR emails

It supports natural input in English and Korean, including mixed-language speech, making documentation easier for multilingual healthcare professionals.

How I built it

I built Nurse Buddy as a React and TypeScript web application using Vite, Tailwind CSS, Supabase, the OpenAI API, and browser-based speech recognition.

I designed the workflow and prompts based on my experience as a registered nurse. The application processes informal clinical observations and transforms them into organized, professional documentation formats.

I also used AI-assisted development tools, including Codex, to help write, review, troubleshoot, and improve portions of the code. I made the final product decisions, tested the application, and refined the documentation output based on real nursing workflows.

Challenges I ran into

One of the biggest challenges was generating documentation that is concise and professional without adding information that the nurse did not provide.

I also worked on supporting mixed Korean and English input, maintaining consistent terminology, and producing different documentation formats from the same clinical information.

Another important challenge was designing Nurse Buddy as a documentation support tool rather than a replacement for clinical judgment. The nurse must review and approve all generated content before using it.

Accomplishments that I am proud of

I am proud that Nurse Buddy was created from a real problem I experience as a working registered nurse.

The project:

Supports multilingual and mixed-language input
Generates multiple nursing documentation formats
Identifies missing information before documentation is generated
Reflects realistic nursing workflows
Combines my nursing experience with my background in computer science

What I learned

Through this project, I learned how to combine clinical workflow design, prompt engineering, frontend development, and AI safety considerations.

I also learned that healthcare AI must be useful without overstating its abilities. Good clinical documentation assistance should improve structure and efficiency while keeping the nurse responsible for verification and final clinical judgment.