Fifteen orchestration engines and MLOps platforms for building, scheduling and running machine learning pipelines at scale — every link below was checked live and is working.
Orchestrators built directly on top of Kubernetes primitives for containerized pipeline execution.
An open-source, container-native workflow engine for orchestrating parallel jobs on Kubernetes, part of the broader Argo Project alongside CD, Rollouts and Events.
Visit official site →A unified Python interface for constructing and managing workflows across engines such as Argo Workflows, Tekton and Apache Airflow, adopted internally at Ant Group.
Visit official site →Simplifies deploying Jupyter Notebooks as Kubeflow Pipelines workflows directly from a notebook, with a v2 release shipping in 2026 for current Kubeflow Pipelines.
Visit official site →Python-first schedulers for data and ML pipelines, usable with or without Kubernetes.
An AI orchestration platform written in pure Python for dynamic, resilient workflows at scale, with 80M+ downloads and adoption at Fortune 500 companies.
Visit official site →An asset-centric data orchestrator for ML, analytics and ETL that tracks lineage and data health across a pipeline rather than just task completion.
Visit official site →Turns any Python function into an observable workflow with a single decorator, backed by an open-source core plus managed Prefect Cloud for production.
Visit official site →Spotify's Python package for building complex pipelines of batch jobs, handling dependency resolution, workflow management and failure recovery.
Visit official site →Frameworks that manage the full ML lifecycle: data prep, training, tuning, tracking and deployment.
An open-source AI orchestration framework managing ML and generative AI applications across their lifecycle, also available as a managed service on the Iguazio platform.
Visit official site →An extensible open-source MLOps framework for reproducible pipelines, now positioned as a unified platform spanning classical ML, LLM pipelines and AI agents.
Visit official site →A Python framework hosted by the LF AI & Data Foundation for creating reproducible, maintainable and modular data engineering and data science code.
Visit official site →Originally built at Netflix, Metaflow makes it quick to build and manage real-life ML, AI and data science projects from a laptop to the cloud without code changes.
Visit official site →An MLOps platform for AI products that combine LLMs and specialized models, automating everything from dataset versioning to pipeline execution and deployment.
Visit official site →Runs with all popular deep learning frameworks and machine learning libraries on Kubernetes, covering tracking, orchestration, optimization and model management.
Visit official site →Manages ML and generative AI workflows end-to-end on the Iguazio Data Science Platform, built on top of the open-source MLRun framework.
Visit official site →Purpose-built for taking trained models into production inference at scale.
An inference platform for deploying any model anywhere, combining developer-friendly packaging with tailored optimization, autoscaling and streamlined operations.
Visit official site →