MLOps Platform / Workflow Orchestrator
6.9k 2026-04-13
flyteorg/flyte
Dynamic, resilient open-source orchestrator for building scalable and reproducible data and ML pipelines on Kubernetes.
Core Features
Scalable and reproducible data/ML pipelines
Leverages Kubernetes for underlying platform
Python SDK for pipeline construction (and other languages)
Deployment on cloud and on-premises environments
Distributed processing and efficient resource utilization
Quick Start
pip install flytekitDetailed Introduction
Flyte is an open-source workflow orchestration platform designed for machine learning and data pipelines. It enables engineers and data scientists to build, deploy, and manage production-grade AI workflows with a focus on scalability, reproducibility, and efficient resource utilization. By leveraging Kubernetes, Flyte provides a robust and distributed execution environment, supporting both cloud and on-premises deployments. Its Python SDK simplifies pipeline definition, allowing teams to coordinate data, models, and compute seamlessly.