With Machine Learning Model Operationalization Management (MLOps), we want to provide an end-to-end machine learning development process to design, build and manage reproducible, testable, and evolvable ML-powered software.
Being an emerging field, MLOps is rapidly gaining momentum amongst Data Scientists, ML Engineers and AI enthusiasts. Following this trend, the Continuous Delivery Foundation SIG MLOps differentiates the ML models management from traditional software engineering and suggests the following MLOps capabilities:
- MLOps aims to unify the release cycle for machine learning and software application release.
- MLOps enables automated testing of machine learning artifacts (e.g. data validation, ML model testing, and ML model integration testing)
- MLOps enables the application of agile principles to machine learning projects.
- MLOps enables supporting machine learning models and datasets to build these models as first-class citizens within CI/CD systems.
- MLOps reduces technical debt across machine learning models.
- MLOps must be a language-, framework-, platform-, and infrastructure-agnostic practice.
This site describes the practices and the current state in MLOps as a quickly evolving field of machine learning, software engineering, and operations. Please note, we expect a basic familiarity with machine learning concepts. For more resources to learn about machine learning, please refer to our References page.
Motivation for MLOps
You will learn for what to use Machine Learning, about various scenarios of change that need to be managed and the iterative nature of ML-based software development. Finally, we provide the MLOps definition and show the evolution of MLOps. more…
Designing ML-powered Software
This part is devoted to one of the most important phase in any software project - understanding the business problem and requirements. As these equally apply to ML-based software you need to make sure to have a good understanding before setting out designing things. We will also discuss when not to use machine learning. more…
End-to-End ML Workflow Lifecycle
In this section, we provide a high-level overview of a typical workflow for machine learning-based software development. more…
Three Levels of ML-based Software
You will learn about three core elements of ML-based software - Data, ML models, and Code. In particular, we will talk about
- Data Engineering Pipelines
- ML Pipelines and ML workflows.
- Model Serving Patterns and Deployment Strategies
In this part, we describe principles and established practices to test, deploy, manage, and monitor ML models in production. more…
State of MLOps
This part presents an overview of software tools and frameworks that manage ML artifacts and cover the whole machine learning cycle. more…
Model Governance | Ethics | Responsible AI (coming soon)
With the recent advances in machine learning and AI, these technologies are increasingly being applied in software systems that interact with people in their daily lives. This progress also led to new questions about how to build security, privacy, fairness, and interpretability, privacy, and security into these ML systems.
Use-cases and Demos (coming soon)
Get a curated list of references ranging from the core MLOps to the Economics of ML/AI. more…