ConnectCare

The "Why": It Started with a Scare

Alright, so the story behind ConnectCare isn't just some random "app idea." It's personal. A little while back, a family member of mine was recovering from a major surgery. Everything seemed fine, but then things took a bad turn, fast. It was all because of a symptom that seemed minor to us, but was actually a huge red flag.

The doctors told us that if we had caught it just two days earlier, the whole emergency could have been avoided. That hit me hard. I kept thinking, "How many other people go home after surgery and have no idea what they're supposed to be looking out for?"

As a problem-solver (that's what we programmers are, right?), I couldn't let it go. I looked around and saw this was a massive, silent problem. That's when the seed for ConnectCare was planted. I wanted to build something that could act as a safety net, a digital lookout for patients and a vital data source for doctors.

The "How": A Python Dev in a JavaScript World

Here's the interesting part. My background is all Python. I think in terms of dictionaries, lists, and simple, clean syntax. The world of JavaScript, with its curly braces and asynchronous-everything, felt like a different planet. And React Native? I knew basically nothing.

So, I didn't code this app in the traditional sense. I built the entire thing using bolt.new .

Think of it like this: it was like having a senior React Native developer as my pair programmer, 24/7. I was the architect. I knew the logic I wanted—the if/else conditions, the data I needed to store, the API calls I needed to make. I would describe these requirements in plain English, and the AI would translate my "Pythonic" thinking into functional React Native code. I was focused on the "what," while the AI handled the "how."

The "Struggle": Learning a New Language (of Prompting)

It wasn't a walk in the park. My first challenge was just getting my head around the basic concepts of React Native. As a Python guy, things like state management, props, and the component lifecycle felt foreign. I had to do a crash course just to be able to describe what I wanted in a way the AI could understand.

The real challenge, though, was learning how to prompt. It's a skill in itself, almost like writing good specs for a function. My first few prompts were too vague, and I got messy results. It's the classic "garbage in, garbage out" principle, but applied to AI.

I had to learn to break down every feature into tiny, specific tasks. Instead of saying, "Build a login page," I had to say, "Create a screen with two input fields for email and password, a 'Sign In' button, and a 'Forgot Password' link. When the button is pressed, validate the inputs and call this specific Supabase authentication function." It forced me to be incredibly clear and structured in my thinking.

The "Aha!" Moment: More Than Just an App

Building ConnectCare taught me a ton. I learned that you don't need to be an expert in every language to build something meaningful. More importantly, it shifted my perspective from being just a "coder" to being a "product builder." My primary tool wasn't an IDE, but my ability to clearly articulate a vision and guide a powerful AI to execute it.

This project is more than just lines of code to me. It's the answer to a question that started with a family emergency, and it's my first real step towards building a startup that I hope can help many others.

Built With

Share this project:

Updates

Submission history