Polish & Pixel | Laravel Cloud, Forge and Deployment Engineering

Where a Laravel application runs is an engineering decision with long consequences: it sets your deployment workflow, your scaling ceiling, your compliance story, and a large part of your monthly bill. The 2026 options are better than they have ever been, and more numerous. Polish & Pixel helps you choose, builds the deployment pipeline, and migrates you off whatever you are on today.

The 2026 options

Laravel Cloud. The first-party managed platform. Push to deploy, managed PostgreSQL and MySQL, Redis-compatible caches, object storage, autoscaling, and scale-to-zero for preview and low-traffic environments. The fastest route from repository to production for most teams, with no servers to patch.

Laravel Forge. Managed provisioning and deployment for servers you own on AWS, DigitalOcean, Hetzner, Vultr, or any VPS. You get the control of your own infrastructure (and its cost profile) with Forge handling PHP, Nginx, databases, queues, SSL, scheduled tasks, and deployments.

Laravel Vapor. Serverless deployment on AWS Lambda. Pay per request, effectively unlimited horizontal scale, and no servers, at the cost of Lambda's constraints on request duration, cold starts, and local disk.

Containers and Kubernetes. For organisations that already run a platform team and want Laravel to fit the same pipelines as everything else. More to build and operate, but the right answer when policy requires it.

Choosing between them

We make the recommendation on four axes:

Team size and ops appetite. No dedicated ops person points to Laravel Cloud. A team that already manages servers and wants cost control points to Forge.
Compliance and data residency. Requirements to keep data in a specific account, region, or network usually mean Forge on your own cloud account, Vapor in your AWS organisation, or containers on your platform.
Traffic shape. Steady traffic favours right-sized servers. Spiky or unpredictable traffic favours Vapor or Cloud autoscaling. Long-running requests, websockets, and heavy background processing steer away from serverless.
Budget and growth. We model twelve months of cost on each option against your actual request and job volumes rather than list prices.

Octane for high-throughput applications

Laravel Octane serves your application from an always-booted process using FrankenPHP, Swoole, or RoadRunner, removing framework bootstrap from every request. For API-heavy or high-concurrency applications it can be the difference between scaling out and staying on one server. We assess whether your codebase is Octane-safe (static state, singletons holding request data, memory growth), make it so, benchmark before and after, and wire Octane into Cloud, Forge, or your container image.

Observability with Nightwatch and Pulse

A deployment is not finished until you can see it. We configure Laravel Nightwatch, the first-party APM, for request, query, job, exception, and outgoing-HTTP visibility, and Laravel Pulse for an at-a-glance health dashboard. Alerts go to the channels your team already watches, with thresholds tuned to your traffic rather than defaults.

Migration engagements

The most common engagement is moving an application off legacy hosting: cPanel shared hosting, a hand-built VPS nobody wants to touch, or a container setup that has outgrown the team maintaining it. A typical migration runs:

Inventory of the current environment: PHP version, extensions, cron jobs, queue workers, storage, environment variables, and every undocumented dependency.
Target build on Cloud, Forge, Vapor, or containers, including queues, scheduler, storage (usually to S3-compatible object storage), and TLS.
Data migration with a rehearsed cutover plan, a measured downtime window (often zero), and a rollback path.
Pipeline with automated tests on every pull request and deployments you can trust.
Monitoring with Nightwatch and Pulse before the DNS change, not after.

Deployment engineering for existing setups

You do not have to migrate to benefit. We also fix deployment pipelines in place: zero-downtime releases, database migration safety, queue worker restarts, horizontal scaling, CDN and caching layers, and secrets management.

Need a hosting recommendation backed by numbers, or a migration executed without drama? Contact us and tell us what you are running today.

Back to services