Ellen Britt, PA, Ed.D., PHom and I had planned to go live on Substack to share the story of how she went from ER PA to vibe coder (it’s a good one, and we were excited to walk you through it in real time).
But… technology had other plans.
After a few attempts to get things working smoothly, we moved over to Zoom so we could deliver the conversation without more delays.
We talked about how Ellen’s clinical background shapes her approach to building with AI, what “vibe coding” actually looks like in real life (not the hype version), and how non-coders can start building in a way that actually feels doable.
Show notes are below…
If you prefer audio:
Ellen’s Story
My conversation with Ellen Britt traces her path from working for nearly two decades in emergency medicine to building AI-powered applications without a technical background. Her experience in the ER shaped how she approaches AI: making decisions with incomplete information, relying on structured thinking, and iterating quickly rather than waiting for perfect clarity.
A key shift came when she stopped chasing the “perfect prompt” and instead focused on problem-solving. By applying the clinical SOAP framework—Subjective, Objective, Assessment, Plan—she was able to break down complex AI-building tasks into manageable steps and make steady progress.
Rather than using done-for-you platforms, Ellen chose to build locally with AI guidance, which gave her a deeper understanding of how things work. This process not only resulted in a functional app but also built confidence and opened up new possibilities for what she could create.
The broader takeaway is that AI is not about tools or prompts alone. It’s about how you think. When you bring structure, clarity, and iteration into the process, AI becomes a powerful partner rather than a source of frustration.
Key Takeaways
1. AI amplifies your thinking, not your prompts
The biggest shift was moving away from trying to “get the prompt right” and focusing instead on clear problem definition. Vague input leads to inconsistent output, while structured thinking produces better results.
2. Incomplete information is not a blocker
Ellen’s ER background trained her to act without having the full picture. That same mindset applies to AI building. You start with what you know, test, and refine as new information appears.
3. The SOAP framework simplifies complex work
Breaking problems into Subjective, Objective, Assessment, and Plan creates a repeatable process. It keeps you grounded, reduces overwhelm, and gives you a clear next step every time.
4. Start small and build in layers
The first version of the app was intentionally simple. No extra features, just something functional. From there, improvements were added step by step. This prevented overload and made progress visible.
5. Iteration beats perfection
Nothing worked perfectly the first time. Progress came from cycles of testing, diagnosing issues, and refining. Each loop improved both the app and her understanding.
6. Treat AI like a collaborator, not a tool
Instead of issuing commands, Ellen had ongoing conversations with AI. She asked questions, challenged responses, and guided the process. That back-and-forth made the work more effective.
7. Control improves confidence
Building locally, rather than relying on platforms, gave her ownership of the code and a clearer view of how everything worked. That visibility reduced intimidation and increased capability.
8. Constraints improve results
Limiting AI to one step at a time and asking for complete code updates (instead of partial edits) reduced errors and confusion. Clear instructions produced cleaner outcomes.
9. Real value is in useful tools, not more content
The conversation highlights a shift from creating content to building tools people can use repeatedly. Practical, problem-solving apps have more value than static resources.
10. The biggest outcome is confidence, not the app
Building something from scratch changed how Ellen sees herself. The process removed the feeling of being “locked out” of AI and replaced it with curiosity about what else is possible.
The Next Step…
Ellen Britt and I are putting together a small beta group of people interested in learning to build AI-powered apps from scratch. We’re working on the details, and if you’re interested in learning more, get on our “first to know” list here. No obligation, just info so you can see if it’s something you’d like to learn.
Recommended Reading
Got questions about building apps with AI? Ask them here ↴
