Build Apps with AI for Healthcare Professionals
A practical introduction to building with AI coding agents, even if you have never written code. You learn to give an agent clear instructions, follow the changes it makes, and check whether the result works. Along the way you build a browser-based clinical calculator and a Python assistant that searches PubMed, then put a project online at a real web address.
What you will learn
- Direct an AI coding agent (Claude Code) to build working software for you, without writing a line of code yourself
- Build a clinical calculator (CHA2DS2-VASc and Cockcroft-Gault) that runs in your browser, then break it on purpose to check it fails safely
- Write a Python script that pulls real data from a free API, and read the JSON that comes back
- Deploy your tool from GitHub to a live public URL, and run a safety pre-flight before you send the link to a colleague
- Run a research project end to end with AI agents: a proposal, a Word document, an extraction spreadsheet, and a second pass that catches what the first missed
- Explain what a language model, an AI agent, a context window, an API, an API key and MCP actually are, using analogies from the clinic
- Keep patient data and API keys safe: know what leaves your laptop, what never goes to GitHub, and why a key is treated like PHI
Course outline
- The tools you need
- The machine: what code, language models and AI agents actually are
- Your first project: a research review with no code at all
- APIs and keys: how your agent reaches the outside world
- Build your first app: a clinical calculator
- The back end: a tool that talks to a real AI model
- Save and share your work with Git and GitHub
- Put it online: deploy your tool to a real web address
Who it is for
- Physicians, nurses, pharmacists, residents, fellows and students who have never written code and want to build their own tools
- Clinicians who have used AI in a chat window and want to go further, with real files, real data and a live tool
- Educators, leaders and administrators who want to judge AI tools and the people who build them
Not for experienced software developers: the course teaches directing, not syntax. All examples are for education and use fictional information. The projects are learning exercises, not validated clinical tools, and should never be used with patient data or for decisions about real patients.