CACAO

CACAO (Cloud Automation and Continuous Analysis Orchestration) is an open-source, cloud-native platform developed and maintained by CyVerse — the NSF research project that created Atmosphere. It is designed to help researchers, educators, and research software engineers (RSEs) deploy software and infrastructure across multiple cloud environments using reusable, community-driven recipes. CACAO addresses the growing complexity of modern cloud service providers by abstracting cloud infrastructure through infrastructure-as-code (IaC) and domain-specific language (DSL) approaches, making cloud resources more accessible and productive for scientific and educational use.

CACAO is built around a core mission of open science: adhering to FAIR (Findable, Accessible, Interoperable, and Reusable) data principles, broadening access to cloud resources for research and education, facilitating collaboration between RSEs, researchers, and educators, and promoting community-driven sharing of best practices for creating and reusing cloud infrastructure recipes.

Key Features and Capabilities

• Integrates and deploys cloud resources into NSF Jetstream2 and AWS, with planned future support for GCP and Azure Terraform.
• Enables users to adapt cloud infrastructure to their specific needs using recipes written for Terraform, Ansible, or Kubernetes — no prior recipe-writing experience required, as community-contributed recipes are available.
• Planned support for non-Terraform-based recipes, including Nextflow workflows.
• Supports Continuous Analysis capabilities, allowing users to automatically activate cloud resources when data or workflow code changes, and shut down resources after workflow execution to optimise cost and efficiency.

Existing Supported Recipes

• Zero-to-JupyterHub
• Kubernetes on demand
• Docker and Docker Swarm
• Magic Castle (from Digital Research Alliance)
• Virtual machines for workshops and classes
• Text-generation-webui
• Danswer
• And many additional community recipes

Documentation and Getting Started

• A comprehensive CACAO Guide is available for new and experienced users.
• Quickstart resources are provided to help users get up and running rapidly.
• A full Manual covers in-depth platform usage and configuration.

CACAO is accessible via the CyVerse platform and currently launches through the Jetstream2 site. As an open-source project, it is positioned to grow through community contributions, supporting a multi-cloud research and education ecosystem aligned with open science standards.