Building a Managed Platform for PHP
For the past year, I have been quietly building a managed hosting platform for PHP applications. Auto-scaling, metrics-driven infrastructure, and zero-config deployments. Here is what I have learned so far.
I've been hinting at this in various posts, but I haven't talked about the bigger picture yet. For the past year, alongside the open source work on Init, FPM Exporter, and System Metrics, I've been building a managed hosting platform for PHP applications.
The idea is simple but the execution isn't: a platform where you push code and everything else is handled. Scaling, metrics, health checks, process management, PHP-FPM tuning, queue workers, scheduled tasks. All the things I've been building as individual packages, wired together into a cohesive platform.
Why another hosting platform?
Every PHP team I've worked with over the past 20 years runs into the same problems. They either use shared hosting and outgrow it, or they move to AWS/DigitalOcean and spend months building deployment pipelines, monitoring stacks, and scaling logic. The infrastructure work never ends, and it distracts from building the actual product.
The existing PaaS options are either generic (Heroku, Railway) and don't understand PHP-FPM, or they are expensive and opinionated in ways that don't match how PHP applications actually work in production.
The open source foundation
Everything I've been publishing as open source is a building block for this platform:
Cbox Init is the process manager that runs inside every container. It handles PID 1, PHP-FPM auto-tuning, health checks, and multi-process orchestration.
FPM Exporter provides the metrics that drive scaling decisions. Pool status, OPcache, queue depths.
System Metrics gives container-aware resource readings. Real memory usage, not misleading host metrics.
PHP Base Images are the standardized container images that everything runs on.
Queue Autoscale uses queueing theory to scale workers based on SLA targets instead of arbitrary thresholds.
Each package works standalone, and developers already use them independently. But together, they form the infrastructure layer of something bigger.
What I've built so far
The platform runs on Kubernetes. Each application gets its own namespace with isolated resources. Deployments are triggered by git push. The build system detects the PHP version, framework, and dependencies, then produces an optimized container image.
The interesting part is the metrics-driven infrastructure. Every container runs Cbox Init with FPM Exporter, which feeds Prometheus. The autoscaler reads these metrics and adjusts PHP-FPM workers, queue workers, and horizontal pod scaling in real-time. It isn't just "add more containers when CPU is high." It uses the same Little's Law calculations from Queue Autoscale to predict demand and scale proactively.
The auto-tuning is what I'm most excited about. When you deploy an app, the platform profiles its memory usage per request, calculates optimal PHP-FPM worker counts based on the container memory limit, and adjusts pool settings automatically. No more guessing pm.max_children.
What is next
I'm not ready to announce a launch date. The platform works for my own projects and a few early testers, but there's a long list of things to build before it's ready for other teams. Database provisioning, SSL automation, team management, billing, and the dozen other operational concerns that separate a side project from a real product.
I will keep publishing the individual packages as open source. The platform is the commercial product; the tools are free.
More updates soon. If you are interested in early access, reach out.