I Automated 1,500 Hours of Manual Work. Here Is What Happened After.

Across a year of automation work I clawed back roughly 1,500 hours of manual labor. Data entry, report generation, repetitive form filling, the kind of work that is too important to skip and too dull to do well by hand. I am proud of the engineering. I am less settled about the rest of it, and I want to be honest about that.

The work itself

A lot of it was browser automation with Playwright. There was a healthcare-adjacent workflow where staff were manually moving data between systems that refused to talk to each other. Hours a day, every day, copying fields from one screen to another. It was error-prone because humans are not built for that, and the errors mattered because it was patient-related data.

So I built bots. They logged in, navigated, read the fields, validated them, wrote them to the other system, and logged every step so we had an audit trail. What took a person an afternoon took the bot a few minutes, and the bot did not get tired and skip a row at 4pm on a Friday.

The bot did not get bored. That, more than speed, was the real upgrade. Boredom is where human errors live.

What the client said

The client was thrilled, of course. The numbers were undeniable. Faster, cheaper, fewer mistakes. They wanted more processes automated immediately. From a pure business view it was a clean win, the kind that makes you look like a genius.

The weird feeling

Here is the part I do not see written about much. There was a person whose job was that copying work. Not their whole job, but a real chunk of it. And after the bot, that chunk was gone. I never met them. I just made a thing that quietly removed a few hours of their daily existence from the books.

I told myself the comforting story. Now they can do higher-value work. And sometimes that is true. But sometimes the higher-value work does not materialize, and the hours you saved the company are hours someone used to be paid for. I do not have a clean answer to this. I am suspicious of anyone who does.

The ethical question nobody asks

When I scope an automation project, the question on the table is always efficiency and cost. Never once has a client asked me what happens to the people currently doing the work. Never. It is just not in the conversation. And I, the person building the thing, also do not usually raise it, because it is awkward and it is not what I was hired to ask.

I am not pretending I have the moral high ground. I built the bots. I took the money. But I think the least I can do is not pretend the question is not there:

• Automation that helps a person do their job better is a different thing from automation that erases the job.
• In a country with limited social safety nets, "they can retrain" is not always a real option.
• The efficiency is real. The displacement is also real. Both things are true at once.

Where I landed

I still do automation work. It is some of the most satisfying engineering I do, taking a dumb repetitive process and making it disappear. But I have stopped pretending it is a pure good. It is a tool, and like every tool it cuts in the direction you point it. The hours I saved were real. So was the person on the other end of them. I keep both of those facts in the same hand now, even though it is heavier that way.