KitOps is an open source DevOps tool that packages, versions, and securely manages AI/ML models, datasets, code, and configurations as OCI artifacts.
An open source DevOps tool for packaging and versioning AI/ML models, datasets, code, and configuration into an OCI artifact.
KitOps is designed for AI/ML platform engineering teams to streamline the packaging, versioning, and sharing of AI/ML projects, enabling reproducibility and secure deployment. It is ideal for organizations needing to track, control, and audit AI/ML assets across the project lifecycle, including compliance with regulations like the EU AI Act.
ModelKits are immutable and signable, making them ideal for secure AI/ML asset management and compliance auditing. Best practices include integrating KitOps with existing OCI registries and DevOps pipelines to maximize security and efficiency. KitOps also supports air-gapped environments, important for enterprise security requirements.
Visit the official website https://kitops.org for detailed installation guides
Install Go programming language from https://go.dev/ if not already installed
Clone the repository or download the KitOps CLI from the releases
Use the provided CLI tools to create and manage ModelKits
Configure your OCI container registry credentials for storing ModelKitskitops create <modelkit-name>
Creates a new ModelKit package including models, datasets, code, and configuration.
kitops tag <modelkit> <version>
Tags a ModelKit with a version to track compatible datasets and models.
kitops unpack <modelkit> --selective
Selectively unpacks parts of a ModelKit to save time and storage.
kitops deploy <modelkit>
Creates a runnable container from a ModelKit with a single command.