The MLOps space can be confusing. So we decided to do something about it.

We launched a new open-source website called MLOps.toys – a curated list of the most useful MLOps tools and projects for training orchestration, experiment tracking, data versioning, model serving, model monitoring, and explainability.

a screenshot of MLOps toys website

Why did we create MLOps Toys?

Because every data science and ML engineering team should have the tools they need to create a strong machine learning infrastructure that supports their end-to-end MLOps needs. Uber, Netflix, Airbnb aren’t the only ones who can build their own machine learning platforms. It’s easier than you might think.

Whether it’s for experiment management, data versioning, serving platforms, or another part of the ML pipeline – we all have different needs and there are TONS of tools to support the ML infrastructure.

But many of these solutions overlap with each other, and it’s difficult to know how to start comparing them in order to find your ideal solution.

A meme that talks about the confusing MLOps space

An easier way to find the ML tools you need

MLOps Toys is our attempt to make some order in the MLOps space with a well-categorized list that includes a simple description, highlighted features and benefits of over 30 machine learning tools and projects within the space.

I hope you will find it helpful 😃

Contribute to MLOps Toys

We need your help! The MLOps space is continuously growing. If you think we’ve missed a relevant project, or are missing  some information about a tool, please create a GitHub issue or a pull request.

We’d love any contribution you can give to make this list as useful as possible!

✏️ Contribute on Github