Model Registry for Machine Learning

Organization, Discovery and Collaboration

Curate your models in a central dashboard that facilitates model discovery across all your ML projects

Deployment automation

Track, deploy, and manage models through the development lifecycle and integrate with CI/CD solutions like Heroku

Unifying ML projects with DevOps

Manage model lifecycle through Git commits and tags, unifying ML projects with the best DevOps practices every step of the way

Build your model registry on top of Git

Reuse existing Git infrastructure for managing ML models together with code, data and metrics. With git as your single source of truth, use GitOps for model deployment.

An example showing git tags and its respective stage in DVC Studio dashboard.

Answer questions like,

Which model version is currently in production?
Who created this model version and when?
Which code and dataset was used to train it?
Who put this model into production?
and more.

Manage your ML models through the entire lifecycle, from development to production.

Create versions and assign stages to models to build a timeline of model actions from experimentation to production.

An example in DVC's Studio product. Shows a page where you can view a list of all your models.

Manage the lifecycle of each model as it moves through staging, production and other stages. See at a glance which model versions are in which stage.

Use the interface of your choice

An example using both CLI terminal and DVC Studio dashboard to get info about a project.

Use the interface that you're most comfortable with, whether you're a data scientist who likes APIs, or a manager who prefers a web UI, or a DevOps engineer who works best with the command line interface (CLI) and uses GitOps principles for model deployment.

Build your model registry with DVC's ecosystem of tools

Use DVC tools for even more benefits, like linking versioned data with your code and models.

An example of the DVC ecosystem. The example shows two layers, the bottom layer representing third part tools and the top layer represents DVC products.

Gain additional capabilities around data versioning, pipeline management, experiment tracking, training automation, and more with DVC tools including DVC CLI. DVC's products are modular, so you can pick those components that extend what you already have. Because you work with your existing cloud and DevOps infrastructure, you can avoid vendor lock-in.

Ready to build your model registry?

Reach out to one of our MLOps experts!

Git-backed Model Registry for Machine Learning