A mindset shift that changed how I build software
The most valuable engineers don't ship more code. Impact, not effort, is what compounds.
Early in my career, my value as a software engineer was measured by output.
How much I shipped. How quickly I executed. How reliably I delivered what was asked.
I believe that that's exactly how it should be at the start.
But while I was at Amazon, I learned a lesson that fundamentally changed how I think about engineering work:
As you grow more senior, output becomes expected. Impact becomes the differentiator.
The value shifts from doing the work to deciding which work is worth doing - and influencing systems, teams, and direction in ways that create outsized positive outcomes.
Refactors still matter. Tickets still get closed. But their true value is no longer in volume - it's in the leverage they create.
I've carried that mindset with me ever since.
Today, I'm incredibly intentional about the problems I take on. I'm drawn to work where a well-placed solution doesn't just help one team or one project - but improves the system as a whole.
That's the lens through which Chiku was built.
Chiku is an automated AI merge request reviewer.
Merge requests are one of the final artefacts of a software engineer's work - and one of the most underestimated friction points in software development. Reviews demand deep focus, introduce waiting time, and still leave room for inconsistency and missed issues.
Yet this step is CRITICAL. It's where quality, maintainability, and long-term velocity are either protected - or quietly eroded.
Improving this part of the system creates compounding returns:
- Earlier bug detection
- Faster, more consistent reviews
- Less cognitive load on senior engineers
- Better outcomes without increasing headcount
And that's high-leverage work.
I built and rolled out Chiku alongside my regular responsibilities - planning, building, and supporting it with the same rigour I'd apply to a production system.
Not as a side project. As a deliberate investment in solving a systemic problem.
Seeing its impact across an engineering organisation reinforced something I've come to believe strongly:
The most valuable engineering work rarely looks like heroics. It looks like removing friction, tightening feedback loops, and quietly raising the baseline for everyone.
And there's a broader lesson here - one that applies well beyond software:
Effort is finite. Leverage compounds.
The work that matters most isn't always the loudest or the busiest - it's the work that makes many other people more effective.
If this way of thinking resonates - about leverage, systems, and impact - I'm always happy to exchange notes with others tackling similar problems.
Sometimes that leads to a conversation. Sometimes to collaboration. Either way, those tend to be the most interesting ones.
Sach
Originally published on LinkedIn.