I knew I was doing the work. I just couldn't prove it.

Every company I've worked at had a leveling document. Somewhere in that document there's a grid that says: to be an L3, you need to demonstrate system design across the team's domain, mentor peers, write well-defined test suites, make sound decisions, and communicate effectively across teams.

I read that document dozens of times. I nodded along. Yeah, I do that. I think I do that. I'm pretty sure I do that.

Then my review cycle came around and I sat down to write my self-assessment and I had nothing. Just vibes. I'd scroll through months of PRs trying to remember which ones mattered. I'd think "I helped a lot of people" but couldn't name when or how.

My manager told me I was doing great. But "great" doesn't fill a promotion template.

The gap

You write a PR on Monday, review three on Tuesday, unblock a teammate on Wednesday, ship a feature on Thursday, and by Friday you've already forgotten Monday. By the time the review cycle arrives, months of evidence have dissolved into "I think I did some good PRs."

It’s not that I wasn’t doing the work. I just couldn’t see it.

That frustration is why I'm helping build Pensero. I wanted the thing I wished I'd had every time I stared at a leveling doc and thought: I know I meet these criteria, I just can't prove it.

I run it on my own work now. Here's what I didn't expect.

I was wrong about my own strengths.

I thought I had a decent sense of where I stood. I'd been writing code for years, I got good feedback, I shipped on time. I had a shape in my head, strong here, weaker there, and I'd never questioned it.

The shape was wrong.

Not dramatically. But wrong enough that I had to sit with it. What I assumed were my strongest areas weren't. Something I'd never paid attention to was quietly one of my best.

The uncomfortable part wasn't having weaknesses. I expected that. It was that I'd been so confident about which ones they were. I had this whole narrative, "I'm great at X, I need to work on Y", and it was built from feelings, not evidence. Partly fictional. Months out of date.

I was wrong about which work mattered.

I assumed my impact lived in the big features. The ones I remembered. The ones I could name in a review.

Memory is a terrible filter for this kind of thing. Some of my best work was stuff I'd already forgotten about. And some of the work I was proudest of barely moved the needle. That's a strange thing to confront.

I also assumed I was mostly working on-roadmap. I wasn't. Not as much as I thought. That's the kind of thing you want to know before a promotion conversation, not during one.

I was wrong about what "my work" even meant.

Every leveling doc has a "collaborates effectively" criterion. Every IC I know skips it in their self-assessment. You spend two hours reviewing a teammate's PR, catch a subtle bug, suggest a better architecture. And none of that shows up anywhere except as a green checkmark on their PR.

I assumed that was just the cost of being a good teammate. Invisible by design.

It wasn't invisible. It was a bigger proportion of my output than I realized. I'd been doing it instinctively and never once thought to mention it in a self-assessment, because it didn't feel like my work.

What I didn’t realize at the time: it’s not just that you can’t see your work. It's that you've been trained to discount certain kinds of it. The work that helps others doesn't feel like yours. So you leave it out. And then you wonder why your self-assessment feels thin.

I was wrong about my bad weeks.

There was a stretch where I felt unproductive. I'd written it off. Too slow, too scattered, nothing to show.

Then I looked at my delivery over that period. The bars were there. Not my biggest weeks, but not a crater either. I'd manufactured a crisis from one bad week and let it color everything around it.

Anxiety has a very selective memory. It remembers the week you struggled and forgets the month you didn't.

What Pensero actually does (and what it doesn’t)

Pensero doesn’t judge you. It doesn’t decide promotions. It doesn’t turn your work into a score or a performance label. What it does is much simpler and, honestly, much more uncomfortable: it shows you the full picture of your work as it actually happened. Every PR, every review, every unblock, every piece of collaboration that normally disappears into the noise, it’s all there… Not filtered by memory. Not shaped by how you felt that week. It doesn’t tell you a story. It removes the one you’ve been telling yourself. Once you can see your work clearly, you stop guessing where you stand.

What actually changed

I stopped trusting my memory of my own work. Not because my memory is bad, but because everyone's is. We overweight the recent, forget the steady, discount the time we spend helping others, and fill gaps with anxiety.

I stopped treating the leveling rubric as a judgment and started treating it as a checklist. Not "do I measure up?" but "where am I on each of these?"

And I stopped waiting for someone else to tell me I was doing well. Which is a weird thing to admit, because I still want that. I still check whether my manager noticed. I still feel a little rush when a peer says something kind about my code. That hasn't gone away. But I stopped needing it to know where I stand.

The evidence was always there. I just couldn’t see it, so I assumed it didn’t exist.