I love spontaneous plans.
But please don’t tell me at 4:30pm that we’re going somewhere at 6.
I need time.
To mentally prepare.
To decide what I’m wearing.
To imagine where we’re going.
To imagine what could possibly happen.
To be ready to be spontaneous.
Yes, I know how ridiculous that sounds. 😅
The older I get, the less impressed I am by “cool business ideas.”
I want to know:
Who is currently paying for this problem to disappear?
Because if the answer is “nobody yet, but imagine if…”
I have already lost interest. 😂
The expensive part of a business problem is not always the software.
Sometimes it’s the 17 tiny things people have been doing manually for years because nobody stopped long enough to ask:
“Why are we doing this this way?” 🤔
The weird part about learning software is that sometimes you spend three hours trying to fix something…
and then the problem turns out to be one tiny thing.
A missing environment variable.
A connection string.
A port.
One stupid little typo.
And suddenly everything works.
I don’t know why, but the relief after finding it is always disproportionate.
I’ve started noticing how often businesses use humans as middleware.
Someone receives the information.
Someone copies it somewhere else.
Someone checks it.
Someone calculates something.
Someone sends it to another person.
Someone turns it into a report.
Then everyone wonders why the process is slow.
Sometimes the biggest automation opportunity isn’t a complicated AI system.
It’s the person who has quietly become the API between five different systems.
A lot of software problems don’t start with software.
They start with:
“We’ve always done it this way.”
Then someone spends three hours copying information from one place to another.
Someone else checks it.
Someone sends it to another person.
Someone turns it into a report.
Someone asks why the numbers don’t match.
And eventually somebody says:
“Can we automate this?”
That question is usually where I become interested.
Not because automation is cool.
Because by that point, there is usually a real problem hiding underneath all that repetition.
There is a special kind of workplace conversation that goes:
“Send the report before 8.”
Okay.
Except the data sheet is not with me.
And I am not physically where the data is.
And the report requires the data.
So now we’re having a meeting with reality.
😂
I think this is one of the reasons I like automation.
Sometimes the biggest problem isn’t that people are slow.
Sometimes the process itself makes speed impossible.
I’ve worked around things that measure pressure.
I’ve built things that store data.
I’ve dealt with systems where one wrong number can affect the next calculation.
Now I spend a lot of time thinking about APIs, databases, authentication, deployments and automation.
Sometimes I look at all of that together and think:
How did my life become this particular combination of things? 😂
I genuinely did not plan the plot this way.
One thing I really appreciate about building software in Nigeria is that “works perfectly” is sometimes a very incomplete sentence.
Does it work when the internet disappears?
What happens when power goes out?
What if the person using it has to continue working anyway?
What if the process you’re automating was being done manually long before you arrived?
Those questions have become much more interesting to me than just getting an endpoint to return 200.
Real life is a terrible test environment.
And somehow, it is also the best one.
I think my laptop and I need to have a serious conversation.
My eye has been hurting and my first instinct was:
“Maybe I need those glasses that turn dark in sunlight.”
Not:
“Maybe I should stop staring at this screen.”
😂
Apparently my solution to excessive screen time was going to be… better equipment for looking at the screen.
Tech people are actually unserious.
Why is taking a screenshot sometimes the beginning of a completely unnecessary archive?
You screenshot something because you need it for later.
Then you never delete it.
Three months later you’re scrolling through your camera roll and finding screenshots of things you don’t even remember wanting.
And somehow you still can’t bring yourself to delete them because:
“What if I need it?”
Something I keep relearning as a backend developer:
The code you write is rarely the problem.
It's the assumptions. That the request will have the right fields. That the user will do things in the right order. That the data coming in will be clean.
It never is.
Half of building a good API is deciding what happens when someone does the unexpected thing.
There is a very specific kind of confidence that comes from doing a calculation correctly…
…and then immediately checking it again because you don’t trust yourself. 😂
I have absolutely done the calculation, looked at the number, stared at it for a few seconds, and then gone back through the inputs like:
“Okay, but where did you betray me?”
A few days ago I got an unsolicited offer of up to $58,000 for a private repo to use in AI training.
The requirements: 100+ merged PRs. 20+ source files. 5,000+ LOC. 6+ months of commit history. Real code reviews.
My first thought wasn't "I could make $58k." It was "I don't have a repo that fits this." 😂
But it tells you something: AI companies now value what we used to treat as noise. Real commit history, review threads, the messy path to the final code.
If you get an offer like this, check who is asking and who owns the code before you share anything. I'm not touching it. But it's a good reason to build in the open and keep your history honest.
One thing I find funny about backend development is that sometimes the only evidence that you did a good job is that absolutely nothing happened.
No error.
No incident.
No angry user.
No dramatic Slack message.
Nothing.
You just quietly prevented a problem that nobody will ever know existed.
Very difficult to put that on a portfolio.
“Here is the bug I successfully prevented from happening.”