Skip to content

Welcome

Make MLOps Easy packages the operational lifecycle of a tabular machine learning project into a reusable framework. It combines a modular Python pipeline with a distributed runtime and a batteries-included CLI so you can train, deploy, and monitor models with minimal ceremony.

What you get

  • Composable pipeline – MLOpsPipeline chains configuration, preprocessing, training, deployment, and observability into a single call that you can drop into notebooks, scripts, or services.
  • Distributed runtime – a FastAPI master coordinates long-running workflows across worker agents; the make-mlops-easy CLI drives the experience.
  • Artifact-first deployments – every training run emits a deterministic directory that contains the model, fitted preprocessor, metadata, logs, and optionally an executable prediction helper.
  • Extensibility hooks – registries let you register custom preprocessing steps, training backends, deployment steps, or observability sinks without modifying the core library.
  • Curated tooling – MkDocs documentation, Makefile tasks, examples, and CI pipelines make teams productive from day one.
┌──────────────┐  submit/poll   ┌──────────────┐
│   CLI user   │ ─────────────▶ │   Master API │
└──────────────┘                └──────┬───────┘
                                        │ assign
                                        ▼
                                  ┌──────────────┐
                                  │  Worker pool │
                                  └──────┬───────┘
                                         │ runs
                                         ▼
                                 ┌──────────────┐
                                 │ MLOpsPipeline│
                                 └──────────────┘

Documentation map

  • Getting Started – install the project, launch the runtime, and walk through the CLI (quick start) or dive straight into the pipeline from Python (CLI reference).
  • Pipeline – architecture overviews plus deep dives into preprocessing, training, deployment, and observability. Learn how to reconfigure or extend each stage.
  • Development – set up a contribution environment, understand the automated checks, and explore CI/CD pipelines.

If you are new to the project, start with Quick Start and then explore the Pipeline Architecture to understand how the moving pieces fit together. The examples under examples/ mirror the documentation, so you can follow along with runnable scripts.