sematic-ai/sematic

Health v2 · maintenance only

Adoption not part of the score

1001 stars · 63 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Sematic is an open-source ML pipeline development platform that lets engineers write end-to-end pipelines in pure Python. Pipelines can run locally on a laptop or on Kubernetes/cloud, with type-safe, traceable, reproducible execution and a web dashboard for monitoring and visualizing artifacts.

Use cases

• build ml pipelines in python
• orchestrate model training and data processing jobs
• run ml pipelines locally then on kubernetes
• track and visualize pipeline artifacts
• make ml workflows reproducible and type-safe
• chain spark jobs with pytorch training
• monitor ml pipeline runs in a dashboard

When to choose

• you want to author ML pipelines in pure Python without YAML or DSLs
• you need local-to-cloud parity for pipeline execution
• you want artifact tracking, lineage, and reproducibility out of the box
• you need to mix heterogeneous compute like Spark, GPUs, and CPUs per step

When to avoid

• you need a battle-tested orchestrator with a large community like Airflow or Kubeflow
• your workflows are not ML/data-science oriented
• you need real-time streaming pipelines rather than batch jobs

Facets

framework · maturity active

workflow-automation scheduling monitoring data-science machine-learning developer-tools machine-learning data-science python cross-platform cli mlops ml-pipelines pipeline-orchestration reproducibility lineage-tracking web-dashboard local-to-cloud pipelines devops kubernetes docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="sematic-ai/sematic")