BigHugger
GH Repository · flyteorg

flyte

Dynamic, resilient AI orchestration. Coordinate data, models, and compute as you build AI workflows.

stars
7,491
30-day movement
+5318/day
Related entries
60
Connections
1
agent-appai-development-toolshacktoberfestdeclarativedata-analysismachine-learninggodata-scienceGoorchestration-enginellmpythongolangflyteworkflowgrpcmakedockerai-agentsscaleagenticfine-tuningkubernetesproduction

Flyte is a workflow orchestration engine for coordinating data, models, and compute in AI and machine learning pipelines. It is declarative, runs on Kubernetes, and supports both Python and Go, with an Apache-2.0 license.

Reach for it when you need to define, scale, and run production ML or AI agent workflows reliably on Kubernetes.

Use it to

  • Orchestrate ML training and fine-tuning pipelines
  • Coordinate data processing and model serving workflows
  • Scale LLM and AI agent workloads in production
  • Define declarative data-science pipelines on Kubernetes

For ML engineers and data scientists building production pipelines

Role
agent-app
Language
Go
Licence
Apache-2.0
Forks
886
Open issues
105
Last push
2026-09-15
Latest release
v0.1.0 · 2019-11-22
topicsorchestrationmachine-learningmlopskubernetesworkflowllm