Skip to content

Welcome to FuseML [fju:zɛmɛl]

Open Source Orchestration for Machine Learning

FuseML is an MLOps orchestrator powered by a flexible framework designed for consistent operations and a rich collection of integration formulas (recipes) reflecting real world use cases that help you reduce technical debt and avoid vendor lock-in.

We follow 5 simple principles

  • Flexibility - create and manage dynamic MLOps workflows connecting different AI/ML tools across multiple infrastructure domains

  • Extensibility - leverage FuseML's set of abstractions and extension mechanisms to add support for your favorite AI/ML tools

  • Composability - build complex MLOps workflows for your projects out of composable building blocks implementing a wide range of machine learning functions

  • Collaboration - use MLOps automation and tool integration recipes created in collaboration by all AI/ML team roles - Data Scientists, Data Engineers, and DevOps Engineers

  • GitOps for Machine Learning - apply the key principles that GitOps is based on, such as versioning and the declarative paradigm to concepts specific to machine learning - e.g. models, datasets, workflows - to provide features such as end-to-end artifact versioning and lineage, tracking, reproducibility and reactive automation

Figure adopted from “MLOps: Continuous delivery and automation pipelines in machine learning”

Where to proceed from here

Well, it depends on what you would you like to do:

  • if you want to find out more about this project and the problems it aims to solve, proceed to the About section.
  • if you are looking to learn about FuseML capabilities, we suggest to take the Quick Start.

  • if you want to learn more about FuseML's core concepts, use one of the following resources:

    • the Workflows section covers FuseML workflows and available workflow extensions
    • the Extension Registry section describes in more detail how FuseML integrates with 3rd party AI/ML services
    • the Installer Extensions section details how FuseML can be used to deploy and manage entire AI/ML tool stacks constructed out of 3rd party AI/ML tools

  • if you are evaluating becoming a contributor to this project, go straight to Contributing section.

Demo video

No time for reading? Here's a sneak preview of the installation process of FuseML components.

Where to find us

The team behind FuseML is reachable through:

We don't accept sponsorship requests right now. We welcome all code contributions instead.