Marzo AI can be deployed fully inside your own cloud environment, so your data never leaves your infrastructure.
We currently support self-hosting on Google Cloud Platform (GCP) using your own VPC, domain, and Google Workspace.
At a high level, self-hosting works like this:
Get a self-host license
Select self-host during the purchase process for your account and provide your license key plus access to the detailed deployment guide.
Prepare your environment
A GCP project and VPC where Marzo AI will run
A domain or subdomain for the app (for example:
marzo.yourcompany.com)A Google Workspace tenant for authentication and integrations (Calendar, Meet, Drive, etc.)
Choose your deployment method
Option A – Guided Docker deployer (recommended)
Run a small “deployer” Docker container in your environment
Open the web UI (from
localhostor a bastion host)Follow the step-by-step wizard to provision all required GCP resources (Kubernetes cluster, database, load balancer, DNS, etc.)
When the deployment finishes, your Marzo AI instance will be available at your chosen subdomain.
Option B – Manual Pulumi deployment (advanced)
For teams that want more control, you can use our Pulumi-based infrastructure project
You configure your GCP project, region, DNS zone, OAuth credentials, and secrets
Run
pulumi upto provision the full stack inside your VPCThis option is better if you have an existing infra-as-code pipeline and want to deeply customize the setup.
Complete the in-app setup
Visit your new Marzo AI URL (for example:
https://marzo.yourcompany.com)Sign in with your company Google account — the first user becomes the admin
Paste your license key into the admin panel
Configure email settings (from-name and from-address for Marzo notifications)
Invite additional users and assign admin roles as needed
Once this is done, your team can start using Marzo AI entirely inside your own GCP project and VPC.
In self-hosted mode, all data stays in your environment. You control the network, storage, and access policies