Run RunF9 in your own cloud

One template installs the whole of RunF9 in your account: the app, its PostgreSQL database, your schedules, analysis, documents and the document-search index. Nothing is stored with RunF9, and you choose the AI (your cloud's own, or none). For a client that requires a different cloud, deploy one copy in each.

Preview: the templates are checked automatically (AWS with cfn-lint, Google with terraform validate) and we do the first install with you. Questions: support@hoshmand.ai.

Microsoft Azure

Deploy to Azure
App
Azure Container Apps
Database
Azure Database for PostgreSQL (flexible server, TLS, 14-day backups)
AI
None, or Azure OpenAI in your subscription
  1. Choose Deploy to Azure and sign in to the Azure portal.
  2. Pick a subscription and create a resource group (for example runf9), and a region.
  3. Fill in the database password, session secret and encryption key (random, 32+ characters each; keep them safe). Leave the rest as they are, or pick Azure OpenAI.
  4. Review + create. In about 10 minutes the deployment's Outputs show runf9Url: open it and create your organization.

Template: runf9.ai/deploy/azure/runf9.json

Amazon Web Services

Download the template
App
Amazon ECS on Fargate behind a load balancer
Database
Amazon RDS for PostgreSQL (encrypted, private subnets, 14-day backups)
AI
None, or Amazon Bedrock in your account
  1. Download the template, then in the AWS console open CloudFormation > Create stack > Upload a template file.
  2. Leave the defaults (or choose Bedrock for AI, with a model you enabled in the Bedrock console). For your own domain, give an ACM certificate ARN.
  3. Tick "I acknowledge that AWS CloudFormation might create IAM resources" and create the stack.
  4. In about 15 minutes the stack's Outputs show RunF9Url: open it and create your organization.

Template: runf9.ai/deploy/aws/runf9.yaml

Google Cloud

Open Cloud Shell
App
Cloud Run
Database
Cloud SQL for PostgreSQL (encrypted, point-in-time recovery)
AI
None, or Vertex AI in your project
  1. Open Cloud Shell in your Google Cloud project (billing must be on).
  2. Paste the commands below. Terraform shows what it will create; type yes.
  3. In about 15 minutes it prints runf9_url: open it and create your organization.
mkdir -p runf9 && cd runf9 && curl -fsSLO https://runf9.ai/deploy/gcp/main.tf
terraform init
terraform apply -var project_id=$GOOGLE_CLOUD_PROJECT

Template: runf9.ai/deploy/gcp/main.tf

After it is running

  • The first person to sign up creates the organization and becomes its owner. Then set sign-up to invite-only (the template's signup setting) and invite your team, or connect your company sign-in (SSO).
  • Organization > Where your data lives shows "Your AWS account" (or Azure, or Google Cloud) and the region.
  • Upgrades: deploy the template again with the new image version; database changes are applied automatically and are always additive.
  • Backups: your cloud's automated database backups (14 days by default). Keep the encryption key with them.