My old portfolio had a familiar problem: it looked like a developer site, but it did not feel like my actual work. Too much of it read like a generic dashboard, and too much of the story leaned on older directions instead of what I spend my time doing now.
The current version is much more deliberate. I wanted the first impression to feel closer to industrial systems and product delivery than to a template. That meant stronger typography, a more editorial layout, less card clutter, and copy that speaks directly to machine vision, automation, tooling, and delivery instead of trying to sound like a startup landing page.
The harder part was content. Design changes are easy to spot, but content drift is what really dates a portfolio. I had to strip out older positioning, rewrite the summary, and center the projects that actually represent the current direction: VisionMetrics, client inspection systems, and the broader mix of industrial software and field-aware product work.
That is probably the best test for a portfolio refresh: not whether it looks newer, but whether it finally sounds like the person doing the work.