Karma #
Karma is an experimental open source system for modeling and managing infrastructure as modular, object-oriented components — where each deployment step is shaped by what came before.
Infrastructure as Consequence.
Rather than building environments top-down, Karma encourages config-driven systems that evolve like a graph — each node connected to its lineage, and each change traceable to its origin.
Karma is built on Adage – a configuration-driven deployment framework for modular infrastructure.
Why Karma? #
- Modular, reusable infrastructure components
- Declarative, visual dependency graphs
- Object-oriented modeling across environments
- Persistent system graph stored in Amazon Neptune
- Eventually: drag-and-drop design for system composition
What Karma Does #
Karma builds and maintains a live infrastructure graph by:
- Ingesting configuration from Git and Parameter Store
- Tracking runtime outputs produced by Terraform
- Inferring relationships and dependencies between components
The graph is stored in Amazon Neptune and exposed via API — allowing other tools to explore the system, simulate changes, or request updates. Karma coordinates those changes based on the graph’s structure and dependencies.
Beyond Infrastructure #
Karma’s graph-based design opens the door to runtime introspection, observability, validation pipelines, and machine learning.
Explore the Data Science perspective →
See It in Action #
Want to see how Karma works in practice?
Check out the Demos page for a full walkthrough of a real-world deployment.
Powered by Adage #
Karma is built on top of Adage – a configuration-driven deployment framework that combines Terraform, Parameter Store, and Terragrunt to manage real-world AWS infrastructure.
Status #
This project is in early development.
Follow progress on GitHub →
Or dive deeper into the core theory behind Karma →