more ideas than time

If you’re anything like me, you have far more ideas than time to build them (or in this case, the capability).

I won’t pretend to be an experienced software engineer, but I can respect the fact that there is a difference between code that works and code you are happy to work with. Simple, maintainable, safe, secure, doesn’t explode if you want to change something, that kind of stuff… I suppose that’s why it’s called software engineering. For me it would be like inheriting an incredibly disorganised Excel workbook with inconsistent formatting, extra tabs, and numbers hardcoded, broken formulas all over the place and someone saying “but it works”.

I’m also aware that there’s a lot more work involved in cleaning up a nightmare workbook than building one correctly from the beginning. I think one of the big challenges when inheriting something, is that you don’t know what was going through the person’s head when they made it, why they made certain decisions, how it was intended to work - and in the past, I have solved that problem by aggressively documenting and hiding guidance all through the things I’ve built.

This is the same kind of mindset I’ve tried to bring into using AI in my day to day life. Yes I could just say “make me this” - but what happens down the road when I want to modify it? Update it? Add something new to it? If I don’t understand the how or why behind it, the AI most certainly is not going to either - and this is where I started to think about how I could port a repeatable, sustainable, scaleable process into building with AI.

Over the past few months, I’ve been building and refining a team of agents as well as a strict framework for creating that yes, sacrifices speed in the name of stability.

I would like to make a few things abundantly clear:

• I am not an expert,
• but I am having fun,
• and I have successfully made a few things with it that haven’t broken (yet).

That’s all for now.