Cross Cloud AI Agent

Bluebrick Cross-Cloud AI Agent is still in BETA. This document is aligned with bricks CLI version 1.33.1. ✅

Bluebricks Atomic Infrastructure™ is a powerful abstraction that enables centralized management, enhanced reliability, and high-performance execution.

Introducing Cross-Cloud AI Agent

Cross-Cloud AI Agent is a tuned LLM developed by Bluebricks that effortlessly migrates infrastructure from and to AWS, GCP, Azure, and other cloud vendors.

The Cross-Cloud AI Agent enables organizations to:

• Instantly replicate infrastructure using Bluebricks AI and Atomic Infrastructure™.

• Manage infrastructure updates seamlessly after migration.

• Predict costs for workloads on the new cloud before migration is applied

How Cross-Cloud AI Agent Works

1. The Terraform State file is provided to the Bluebricks Cross-Cloud AI Agent.

2. The agent processes the State file and searches for equivalent Blueprints on GCP.

3. Once a match is found, the agent maps the provided parameters to the new Blueprints and supplies them as input.

Cross-Cloud AI Agent won't initiate a deployment as part of the BETA phase.

How to Use Cross-Cloud AI Agent

1. Ensure that bricks CLI is installed on your local machine by running:

bricks version

If not installed, follow the official Bluebricks installation guide to set it up.

1. Open the terminal and connect to your Bluebricks account by running:

bricks login

1. Generate API Token and set it environment var (use Long Live Tokens article for more info)

export BBX_API_TOKEN=<api_token>

1. Prepare your Terraform State file, then start the Cross-Cloud AI Agent by running the following command:

bricks cross-cloud --file statetests/rds.tfstate.json

Replace <path_to_state_file> with the actual path to your Terraform State file.

Understand the Output

The Cross-Cloud AI Agent generates Terraform files that are compatible with the target cloud. The files consist of equivalent modules and values matching the provided state.

The agent organizes the output files under a root folder called outputs, with a specific path for each discovered module, relative to the working directory.

./outputs/cloud-migration/<module>/*.tf

Each matching module will include a variables.tf that takes the equivalent values as defaults for its matching vars.

Runtime Output Example

See below the output of the Cross-Cloud AI Agent when migrating the S3 bucket from AWS to GCP:

Next Steps

1. Review the generated code and modify it as needed.

2. Create and publish Artifacts for each generated code to your organization’s registry. Follow the How to Create an Artifact guide.

3. Create a new Blueprint, including the required Artifacts published in the previous step. Follow the How to Create a Blueprint guide.

4. Install the created Blueprint using the following command:

bricks bp install --name <blueprint_name> --props ./outputs/gcp_packages/<namespace>/<package_name>/props.json

Replace <namespace> and <package_name> with your Blueprint's namespace and name.

This ensures the installation uses the correct mapped parameters from the migration process. 🚀