Self-hosting

CloudAEye, inside your own cloud. Run CloudAEye Code Review in your own infrastructure, with SSO through Okta or Keycloak over OIDC or SAML 2.0, two-step verification, and your own LLM credentials.

Book a demo All use cases

Your infrastructure. Your identity provider.

Self-host in your own infrastructure with support for SSO/SAML (Okta, Keycloak) and MFA.

CloudAEyeSelf-hosted deployment

What runs in your AWS account

Entry point
Application Load Balancer, the only public ingress
Runtime
Amazon ECS cluster
Private services
MongoDB, ChromaDB, indexing worker
Queues
Amazon SQS for indexing and reviews
Logs and metrics
CloudWatch

How your team signs in

OktaOIDC or SAML 2.0
KeycloakOIDC or SAML 2.0
Email and passwordWith TOTP 2-step verification

Tenant admins can require 2-step verification and set a password age from 30 to 365 days.

Enterprise SSO providers keep enforcing their own verification policies.

Review code without
sending it anywhere.

Some teams cannot let code leave their network. Self-hosted CloudAEye runs in your own AWS account, behind your identity provider, with the model you choose.

Data stays in your VPC.

Choose self-hosting when your data must not leave your VPC. The databases and workers are private, and the load balancer is the only way in.

  • DatabasesPrivate, not exposed
  • IngressOne load balancer
  • AccessRestricted by CIDR
Read the docs

Bring your own LLM.

Provide your own LLM credentials, including Amazon Bedrock models.

  • Your credentialsSupported providers
  • Amazon BedrockSupported
  • ManagedCloudAEye connectivity
Read the deployment guide

Enterprise sign-in.

Connect Okta or Keycloak over OIDC or SAML 2.0. Password users can turn on 2-step verification, and admins can require it.

  • SSOOkta, Keycloak
  • ProtocolsOIDC, SAML 2.0
  • MFATOTP 2-step verification
Read the Okta guide

The whole platform, self-hosted.

Run the CloudAEye console, reviews, the MCP server for coding agents, a GitHub App, and a Jira Forge app in your environment.

  • ReviewsPull requests and pre-commit
  • Coding agentsSelf-hosted MCP
  • IntegrationsGitHub App, Jira Forge app
See the architecture

Deploy with Terraform.

The CloudAEye Terraform bundle provisions the AWS infrastructure and deploys CloudAEye's containers into your account.

Read the deployment guide
  1. Set your configuration in terraform.tfvars.Your LLM provider and model, and the CIDR ranges allowed to reach the app.
  2. Run Terraform to provision Amazon ECS and deploy CloudAEye.
  3. Connect your identity provider.Okta or Keycloak, over OIDC or SAML 2.0.
  4. Set up the self-hosted GitHub App and MCP server for your developers.