From Overthinking to Cloud Coding: My SaaS Development Journey

Domain purchased ✅ Next stop: Building my first SaaS product

You know that feeling when you finally hit "purchase" on a domain name you've been eyeing? That's exactly what happened when I secured mrvedmutha.com. After vibe coding my entire website (which I covered in my last blog), I'm now ready to dive into the exciting world of SaaS development.

But before you start expecting drumrolls and a grand reveal of my groundbreaking idea—sorry to disappoint, that's not happening in this post. Instead, let me take you through the fascinating process and crucial decisions I've made over the past few days before officially starting work on my SaaS application.

The Perfectionist's Dilemma

Here's the thing about me: I overthink everything. Every single instance of my life, every task, every decision—I'm constantly seeking perfection. This trait followed me right into my startup journey. Before even having a concrete SaaS idea, I spent an entire week (maybe two) agonizing over the perfect tech stack for my future product.

Then I stumbled upon a game-changing video by Hitesh Choudhary. In it, he shared a profound insight that completely shifted my perspective. He emphasized that no tech stack is perfect—the key is being confident in what you know and simply executing the process. While he was discussing this in the context of third-party authentication tools, his message resonated far beyond that specific use case. It was about embracing action over analysis paralysis.

Fast forward to today: I've stopped obsessing over the perfect tech stack. Instead, my focus has shifted to something equally important—code management and productivity optimization.

The Multi-Machine Development Challenge

My development workflow spans two different machines, each serving a specific purpose in my daily routine. During the day, I work on my Mac Mini whenever I find free time, squeezing in coding sessions between other commitments. When I'm home in the evenings, I switch to my Windows laptop for more focused development work.

This setup, while flexible, introduced two major headaches:

Compatibility Issues: Making git commits on my Mac system and then switching to my Windows laptop created frustrating inconsistencies. I'd often find myself dealing with the same commit states from the previous night, requiring careful synchronization to ensure my GitHub repository reflected the latest changes.

Storage Space Nightmares: The second problem was even more pressing. Working across multiple projects meant constantly managing the dreaded node_modules folder—that package directory containing all your development dependencies required to run and test your website locally.

Here's where things get ridiculous: while a typical website project might be around 500MB, the actual disk space consumption balloons to nearly 50GB+ for just one project. This happens because of the countless subfolders that clusters need to work efficiently, creating a storage management nightmare.

Exploring Cloud-Based Solutions

Determined to solve this workflow friction, I began researching cloud-based development environments. My first discovery was CodeSandbox.io, a platform that deploys mini virtual machines where you can code directly in your browser or connect with your favorite coding tools like Cursor.

The initial experience was promising—CodeSandbox worked exactly as advertised. However, the pricing model gave me pause. The free tier provides 400 credits, with 5 credits consumed per hour of usage. Given my 4-5 hour daily coding sessions, I could only use CodeSandbox for 10-15 days with free credits. The paid option offered 100 hours for $12, with additional usage charged at $0.015 per hour.

While functional, I wasn't convinced that paying for a pro version just to code was the right move, especially since I'd been successfully managing projects and my portfolio website files locally on my personal SSD.

The Search for Better Alternatives

My research led me to Replit, which appeared to offer more value. Replit comes with an AI code assistant similar to Cursor and provides 100 agent work credits for $20 per month, while also handling CI/CD operations. On paper, it seemed like a sweet deal.

But again, I found myself hesitant. The agent credits felt insufficient for my usage patterns, and jumping to the next tier would double my monthly cost to $40—an investment I wasn't ready to make for what I considered "vibe coding."

The Cloud Server Solution

Then I discovered what turned out to be the perfect alternative: purchasing my own cloud server and hosting a code server via SSH. This approach required a bit more technical setup, but the benefits were substantial.

I purchased a server from Hetzner for around $5 per month. While the initial installation involved a learning curve, ChatGPT proved invaluable in simplifying the process. The setup requires SSH keys, but these are surprisingly easy to configure once you understand the basics.

The Sweet Spot: Power Meets Affordability

Now I have the best of both worlds: a cloud coding service with AI assistance (through Cursor) for just $25 total monthly cost—$5 for the Hetzner server plus $20 for Cursor's AI features. This solution offers better security (thanks to SSH key authentication), seamless cross-device compatibility, and unlimited usage without worrying about credit consumption.

The result? I can code better, safer, and more efficiently across all my devices without the storage and compatibility headaches that were slowing me down.

Looking Ahead

This journey from overthinking tech stacks to solving real workflow problems has taught me something valuable: sometimes the best solutions aren't the most popular or heavily marketed ones. Sometimes they require a bit of DIY spirit and willingness to learn new tools.

If you're interested in a detailed tutorial on setting up this cloud server coding environment, let me know! I'm considering creating a comprehensive walkthrough to help others who might be facing similar development workflow challenges.

For now, I'm excited to focus on what really matters: building my SaaS product with a streamlined, efficient development setup that actually supports my productivity rather than hindering it.

Next up: Actually building the SaaS (and maybe, just maybe, revealing what it is).