I’ve read a lot of stories about solo founders: the late hours, the brainstorming, the coding, the debugging, and then eventually the launch. But I want to reframe what it means to be a solo founder, especially in this age of AI. It isn’t your traditional bootstrapping, so I’m going to share my own experience doing it.
The Idea Book That Collected Dust
I’ve had an idea book for years. Every now and then I’d flip through it, taking stock of what I knew how to do versus what I’d need help with. And honestly? Looking at that list was discouraging.
When you’re bootstrapping or as I like to say, holding things together with socks and scotch tape you don’t have the resources to hire everyone you need. And let’s be real: not all of us have plug-and-play friends who happen to have the exact skills we need and are willing to work for free or equity.
So where did that leave my idea book? Sitting on a shelf while I worked a full-time job, without the time or resources to learn the skills I needed to finish those projects.
AI Entered the Chat (But I Didn’t Reach for the Book)
Even when AI burst onto the scene, I still didn’t pull that idea book down. Instead, I used AI to fuel my current career, to innovate, automate, and build for someone else’s company.
Then life happened. I found myself on the outside looking in, and my first instinct was to just find another job. I still wasn’t reaching for the idea book. But as the pool of laid-off professionals grew and the job search dragged on, I wanted something else to fill me. Because applying for endless opportunities wasn’t doing it.
So I finally pulled down the idea book.
Some Ideas Had Expired. Others Hadn’t.
I’ll admit, some of those ideas? The window had closed. Someone else had already built them, or the moment had passed. But I started thinking about the present instead of the past. What had I seen in my most recent role? Where were the gaps? Where was there still a need?
I went back through the book and made a decision: let’s launch one of these things. Let’s find something interesting to me that might be interesting to other people too.
Following the Data
When I think about technology, I’ve always had a love for data. Not necessarily the coding part, but the results. Still, I knew I needed to understand how to retrieve and present it. So I took a deep dive into APIs and endpoints, learning how automation could pull data and how AI could help tell different versions of the same story.
Recently, I’ve become fascinated with climate and how differently people around the globe experience weather. We’re all living on the same planet, but having very different experiences. I thought: wouldn’t it be great to get a view of what that looks like, even if just through data?
I looked through my library of domains and found one I’d been sitting on: twoclicks.ai. The idea was simple – how quickly can I retrieve useful information? If I could get it in two clicks, that would be great.
So I started looking for resources to aggregate climate data. And while many of these resources were free, the time and effort it took to locate them and extract what I needed was more than I expected. That’s when I decided to build a platform that brought the top data points together in a way that made sense.
From Solo Founder to AI-Supported Founder
This is where the shift happened: from solo founder to AI-supported founder.
I knew what I wanted to build, but I still had gaps in my skillset. With the help of AI-assisted coding, I was able to build out what I needed. And it paired nicely with what I’d already done even offering suggestions on how to improve it.
I decided to make AI part of my team.
The time it would have taken me to build these ideas solo was cut in half. Every time I start a new project, I ask how long something should reasonably take, and then I beat those odds. It’s not perfect or seamless. Some projects have gotten so tangled that I’ve had to hit the reset button at least twice. But the end result is scalable code that I can move based on demand and interest.
Working Smarter, Not Harder
As my code lives in GitHub, I found myself constantly cutting and pasting, sending things back and forth to my AI tools. That felt inefficient. I wanted to work on things in place.
There are ways to connect AI tools to GitHub, but I wanted something that already lived there. So I added GitHub Copilot, and it’s been amazing. My pull requests used to just say “edit.” Now I get meaningful descriptions of exactly what changed, which is incredibly helpful when I’m troubleshooting across multiple iterations.
Setting the Rules of Engagement
Here’s the thing about AI: if you let it, it will over-engineer a simple problem. It makes things tedious and hard to follow. In Claude’s case, I actually give project instructions to ask me before creating additional documentation or making changes while I’m troubleshooting. I tell it when to build and when to wait.
These instructions have been huge for reducing wasted cycles. I don’t want to stare at my screen watching lines of code get rewritten without any context. When I first started using AI to refine my code, that’s exactly what happened it would immediately start making changes. It also asked repetitive questions about my environment every time something went wrong. So I added another instruction: once something is confirmed, don’t ask again.
This is the essential work you have to do with AI. This isn’t vibe coding. This is actual coding. AI-assisted coding and it’s in a completely different lane. A technical background can be helpful when scaling AI, along with a good strategy to get the most out of it.
Building to Scale Without a DevOps Team
Part of my planning is to explore what tools are best for building MVPs that can scale. I found some great resources that saved me the headache of doing a full infrastructure setup, which would’ve eaten into my development time.
I don’t have a DevOps team. It’s just me. My main goal is: how quickly can I get started? So I use services that let me skip the layers of sysops and get straight to building. It’s been a great experience.
Am I a Solo Founder?
From a human standpoint, yes. But am I solo in the workspace? I’d say no. I have a number of resources that are incredibly helpful.
I’ll continue to leverage these tools because they let me build what I want on my own timeline, whether that’s 11 in the morning or 11 at night.
Will I need a team eventually? Of course. At some point I’ll need to fully embrace the role of founder and look at things from a higher level. I can’t always be head-down in the weeds, because then I won’t be able to see the whole field.
But for now, I’ll stay the course. I’ll keep setting the rules and instructions for how AI works with me. I’ll keep exploring new options.
That’s the benefit of being a startup founder even one held together with socks and scotch tape.
https://github.com/features/copilot
https://mstechdiva.com/database-selection-for-engineering-projects/
Discover more from MsTechDiva
Subscribe to get the latest posts sent to your email.