If you're here expecting advice from an AI expert, you're going to be disappointed. I'm not one.
I'm someone who works in leadership at a SaaS company who caught the AI bug. What started as curiosity turned into building small tools. Then automating pieces of my job. Then staying up far too late because I had one more idea I wanted to see if I could make work.
Somewhere along the way, I realized I wasn't just learning AI. I was teaching myself software engineering. That's what this site is about. I'm documenting the journey of becoming an AI engineer in public: the wins, the mistakes, the projects that somehow work on the first try, and the ones that end with me asking an LLM why everything is on fire.
Why "Ship It & Pray"
The name captures exactly how this journey feels. Every project starts with confidence, slowly descends into confusion, and eventually reaches the point where you deploy it and hope nothing explodes.
Sometimes that's because I forgot to test something. Sometimes it's because I accidentally let the AI write half the application before I fully understood what it was doing. Either way, we're learning.
The code is only 20%
One thing I've learned surprisingly quickly is that writing code is only a fraction of what software engineering actually is.
Learning Python, TypeScript, Go, or whatever language comes next is the fun part. It's tangible. Every day you can point to something new you built. The harder part is everything else:
- Understanding someone else's codebase
- Learning Git without making your coworkers hate you
- Designing systems instead of scripts
- Debugging problems that have nothing to do with the line of code you're staring at
- Testing, deployment, monitoring, security, documentation
- Knowing when to build something, and when not to
That's the part nobody tells you about.
Note: I suspect those things make up 80% of what being a software engineer actually is. The code is just the entry ticket.
What you'll find here
This site is my attempt to learn all of it. You'll find write-ups of side projects, AI experiments, architecture decisions I got right, architecture decisions I absolutely got wrong, prompts that worked, prompts that didn't, and lessons learned while trying to build real software instead of toy demos.
I'm not interested in becoming someone who can make impressive demos. I want to become someone who can build software that people actually rely on.
Come along
If you're on the same journey, whether you're coming from product, support, operations, customer success, or somewhere completely different, I hope you'll find something useful here. Maybe we'll learn from each other's mistakes. Maybe we'll even ship something we're proud of.
Either way, welcome aboard. Let's learn to write code. More importantly, let's learn to think like engineers.
You can read a bit more about me and this site, or subscribe via RSS.