---
title: "Migrating Our Apidemia Website from Dotkernel 1 to Dotkernel Light"
description: "A real case study: how Apidemia migrated its own website off an aging PHP framework onto Dotkernel Light, the same stack it migrates clients onto."
og:
  image: images/app/opengraph/og-migrate-apidemiacom.png
---

# Migrating Our Apidemia Website from Dotkernel 1 to Dotkernel Light

Published on: 17-09-2026

We tell clients to modernize before an aging platform turns into an emergency.
apidemia.com was written in Dotkernel 1 which was based on Zend Framework 1 (ZF 1), a framework that has been out of support for years.
We still have clients and platforms running on ZF 1, but we have applied custom updates and security patches to keep it running smoothly.

Even though it wasn't really a necessity, we decided to take our own advice to run our platforms on modern applications.
Here's what migrating our own site actually involved.

## The problem

The site itself was running on an old, hand-maintained Dotkernel 1 with ZF 1 at its core, which was fully functional.
It went through years of small fixes and little architectural change, which proves its sturdiness.

Still, it's the same pattern we walk into on client platforms every week.
Content lived scattered across ad-hoc templates, there was no consistent way to expose pages to anything but a browser, and every AI coding tool we pointed at the codebase for content or copy work produced very little of value.
We were telling prospects their old platform couldn't support modern tooling while ours couldn't either.

## The analysis

We audited it the way we audit any client platform: what's actually salvageable, what standing still actually costs, and what "done" looks like.
The verdict was the one we give most businesses running something this old - a full rebuild on a supported, PSR-based stack made more sense than patching a framework with no upgrade path left.
Content needed one consistent pipeline instead of scattered templates, and the site needed to be readable by more than a browser, since a growing share of 2026's traffic is AI agents researching a company before a human ever looks.
On plenty of sites in 2026, AI agents already read more pages than people do, which changes what "readable" has to mean.

## The solution

We rebuilt apidemia.com on [Dotkernel Light](https://github.com/dotkernel/light) - the same Mezzio/PSR-15 stack, on Laminas components, that we migrate clients onto.
One template handles the homepage; everything else runs through a single markdown-driven content pipeline, so every page - services, knowledge base, this article - is both a normal web page and, on request, a clean markdown document any AI tool can read directly, with no scraping involved.
This is the practical shape of what's now called AI SEO (Artificial Intelligence Search Engine Optimization).

We built it the same way we build for clients: incrementally, with Claude Code and equivalent AI agents doing a large share of the implementation, under human review at every step.
If that delivery model works, our own site is the proof, not just the claim on our homepage.

## How can this help you?

If your own site or platform is running on something this old, the audit that got us here is the same one we'd run for you - a clear picture of what's salvageable, what isn't, and what it would actually cost to fix.

## Why work with Apidemia on this?

Twenty years of PHP, and we're an official Laminas commercial vendor - one of a short list endorsed by the technology's own stewards.
Most of our engineers are ZCE (Zend Certified Engineers).
We deliver with AI agents under human review, on our own site as much as yours.

[Talk to us](/#contact) about your platform.