Connect a GitLab or GitHub repository, choose how many instances to start with, set a monthly budget cap. CloudAutoScale builds, runs and scales your application on infrastructure you never have to see.
No Kubernetes. No Proxmox. No cloud console. Just your code and your limits.Everything happens in one panel. You never log in to a server, a cluster or a hypervisor.
Point CloudAutoScale at a GitLab or GitHub project. Add deploy credentials, environment variables and secrets in the panel. We build from your Dockerfile or detect your stack.
Pick the number of instances to start with and the maximum you are willing to spend per month. The cap is a hard ceiling, not a suggestion.
Traffic goes up, instances go up. When workers run out of room we add capacity: bigger nodes, more nodes, or both. Traffic drops, everything scales back down.
Spikes, scale events and budget thresholds trigger email alerts. Restart, roll back or redeploy from the panel whenever you want.
Push to your branch and we build and ship. GitLab (cloud or self-hosted) and GitHub, private repositories included.
Instance count follows real request load measured in minutes, not a schedule you have to guess in advance.
Set a monthly maximum. Scaling never crosses it, and you are warned well before you get close.
Vertical and horizontal. Workers get bigger and more numerous as your application needs it, then shrink again.
Run on our own Proxmox clusters, or attach your physical servers as capacity. Mix them if you like.
Kubernetes runs underneath. You never configure it, see it or debug it. Restarts and rollbacks are one click.
Every application runs in its own isolated namespace on clusters we provision and manage through the Rancher API. A control plane watches request metrics and budget usage, scales pods first, and when pods are not enough, provisions new worker capacity on Proxmox or brings your own servers into the cluster.
Take capacity from us or attach your own servers. Both are managed from the same panel.
Emre runs production Kubernetes, Proxmox and network infrastructure for hosting and SaaS companies. CloudAutoScale is the layer he kept rebuilding for customers by hand: Git in, running application out, cost capped.
CloudAutoScale turns a Git repository into a running, autoscaling application on our own Proxmox clusters or on your servers, without exposing Kubernetes to the customer. Early-access customers are being onboarded now.
No. A Dockerfile helps but is not required for common stacks. You never interact with Kubernetes directly; the panel is the only interface.
Scaling stops at the cap. You receive an email before you reach it and again when you do. Your application keeps running at the capacity the cap allows.
Yes. Physical servers in your data center or office can be attached as worker capacity. We install the agent, join them to the cluster and operate them like any other node.
GitLab, both gitlab.com and self-hosted, and GitHub. Private repositories work through deploy tokens or deploy keys you add in the panel.
Request rate, latency and resource usage per application, measured over a short window so bursts are caught in minutes, not hours.
We are onboarding early-access customers. Request access below and we will get in touch to discuss your workload.
Tell us what you run and where it lives today. We will reply within one business day.
[email protected]