Mine has fifteen repos, three years, and $1,608.35 to show for it.
A repo nobody uses and a paycheck from one company dry up the same way: the day somebody decides you aren't needed.
A graveyard is not a backup plan.
Everything between the idea and the working product is approaching free. The bottleneck moves upstream.
"Here's a rough version. Does this solve the problem?"
That's the product engineer posture.
The ticket-taker posture is "I'll let you know when it's done."
Stop measuring your week in tickets closed. Start measuring it in problems solved.
If you shipped four features and none of them moved a metric, you didn't have a good week. If you shipped nothing but figured out why a feature wasn't landing, you did.
Talk to a user this week. Not through a PM. Not through a survey. A real conversation.
Ask them what they were trying to do the last time your product frustrated them.
Ramp's engineering principles include "build what the customer really needs, not what they ask for" and "build opinionated systems."
Ramp expects engineers to understand the problem deeply enough to have an opinion. Not "implement the ticket." Not "ask the PM." Have an opinion.
The new engineering job isn't writing code. It's making the codebase safe for anyone to contribute to.
In 2024, 80% of my department disappeared. When your identity is tied to a skill and that skill becomes a commodity, it doesn't feel like an opportunity. It feels like a rug pull.
What's replacing it is harder. You have to understand the problem. You have to form opinions. You have to ship things that might fail and then figure out why.

