Trace Id is missing
Skip to main content

Azure Machine Learning

Use an enterprise-grade AI service for the end-to-end machine learning lifecycle.

Build business-critical machine learning models at scale

Azure Machine Learning empowers data scientists and developers to build, deploy, and manage high-quality models faster and with confidence. It accelerates time to value with industry-leading machine learning operations (MLOps), open-source interoperability, and integrated tools. This trusted AI learning platform is designed for responsible AI applications in machine learning.

Video container

Accelerate time to value

Build machine learning models leveraging powerful AI infrastructure and orchestrate AI workflows with prompt flow.

Collaborate and streamline MLOps

Quick ML model deployment, management, and sharing for cross-workspace collaboration and MLOps.

Develop with confidence

Built-in governance, security, and compliance for running machine learning workloads anywhere.

Design responsibly

Responsible AI to build explainable models using data-driven decisions for transparency and accountability.

Watch the webinar Uncover Predictive Insights with Analytics + AI

Support for the end-to-end machine learning lifecycle

Data labeling

Label training data and manage labeling projects.

Data preparation

Use with analytics engines for data exploration and preparation.

Datasets

Access data and create and share datasets.

Back to tabs

Azure Machine Learning for Generative AI

AI workflow orchestration

Simplify the design, evaluation, and deployment of large language model–based applications with prompt flow. Easily track, reproduce, visualize, and improve the prompts and flows across a variety of tools and resources, learn more about Generative AI in Machine Learning.

Managed end-to-end platform

Streamline the entire large language model lifecycle and model management with native MLOps capabilities. Securely run machine learning anywhere with enterprise-grade security. Mitigate model biases and evaluate models with the Responsible AI dashboard.

Flexible tools and frameworks

Build deep-learning models in tools such as Visual Studio Code and Jupyter Notebooks, using flexible frameworks such as PyTorch or TensorFlow. Azure Machine Learning is compatible with ONNX Runtime and DeepSpeed to optimize training and inference.

World-class performance

Use purpose-built AI infrastructure designed to combine the latest NVIDIA GPUs and InfiniBand networking solutions up to 400 Gbps. Scale up to thousands of GPUs within a single cluster with unprecedented scale.

Accelerate time to value with rapid model development

Improve productivity with a unified studio experience. Build, train, and deploy models with Jupyter Notebooks using built-in support for open-source frameworks and libraries. Create models quickly with automated machine learning for tabular, text, and image data. Use Visual Studio Code to go from local to cloud training seamlessly, and autoscale with Azure AI infrastructure, powered by the NVIDIA Quantum-2 InfiniBand platform. Design, compare, evaluate, and deploy your prompts for large language model–based applications with prompt flow.

A screen shot of welcome to the Azure machine learning studio
A screen shot of pipeline_Four screen

Collaborate and streamline model management with MLOps

Streamline the deployment and management of thousands of models in multiple environments using MLOps. Deploy and score ML models faster with fully managed endpoints for batch and real-time predictions. Use repeatable pipelines to automate workflows for continuous integration and continuous delivery (CI/CD). Share and discover machine learning artifacts across multiple teams for cross-workspace collaboration using registries and managed feature store. Continuously monitor model performance metrics, detect data drift, and trigger retraining to improve model performance.

Build enterprise-grade solutions on a hybrid platform

Put security first across the machine learning lifecycle using the built-in data governance in Microsoft Purview. Take advantage of the comprehensive security capabilities spanning identity, data, networking, monitoring, and compliance, all tested and validated by Microsoft. Secure solutions using custom role-based access control, virtual networks, data encryption, private endpoints, and private IP addresses. Train and deploy models anywhere, from on premises to multicloud, to meet data sovereignty requirements. Govern with confidence using built-in policies and compliance with 60 certifications, including FedRAMP High and HIPAA.

A screen shot of microsoft Azure machine learning page
A screen shot of use responsible use of AI practice throughout the life cycle

Use responsible AI practices throughout the lifecycle

Evaluate machine learning models with reproducible and automated workflows to assess model fairness, explainability, error analysis, causal analysis, model performance, and exploratory data analysis. Make real-life interventions with causal analysis in the Responsible AI dashboard and generate a scorecard at deployment time. Contextualize responsible AI metrics for both technical and non-technical audiences to involve stakeholders and streamline compliance review.

Build your machine learning skills with Azure

Learn more about machine learning on Azure and participate in hands-on tutorials with a 30-day learning journey. By the end, you'll be prepared to take the Azure Data Scientist Associate Certification.

A person working on a laptop in a conference room

Key service capabilities for the full machine learning lifecycle

  • Collaborative notebooks

    Launch your notebook in Jupyter Notebook or Visual Studio Code for a rich development experience, including debugging and support for Git source control.

  • Automated machine learning

    Rapidly create accurate models for classification, regression, time-series forecasting, natural language processing tasks, and computer vision tasks with automated machine learning.

  • Drag-and-drop machine learning

    Use machine learning tools such as designer for data transformation, model training, and evaluation, or to easily create and publish machine learning pipelines.

  • Responsible AI

    Build responsible AI solutions with interpretability capabilities. Assess model fairness through disparity metrics and mitigate unfairness.

  • Registries

    Use organization-wide repositories to store and share models, pipelines, components, and datasets across multiple workspaces. Capture lineage and govern data using the audit trail feature.

  • Managed endpoints

    Use managed endpoints to operationalize model deployment and scoring, log metrics, and perform safe model rollouts.

Comprehensive security and compliance, built in

Get started with an Azure free account

1

Start free. Get $200 credit to use within 30 days. While you have your credit, get free amounts of many of our most popular services, plus free amounts of 55+ other services that are always free.

2

After your credit, move to pay as you go to keep building with the same free services. Pay only if you use more than your free monthly amounts.

3

After 12 months, you'll keep getting 55+ always-free services—and still pay only for what you use beyond your free monthly amounts.

Learn how customers are using Azure Machine Learning to innovate with AI

“PyTorch and Azure Machine Learning are the perfect match for our research team goals, saving time to create disruptive innovation.”

Orlando Ribas Fernandes

Co-Founder and CEO, Fashable

“Our teams usually test [data], get results, and then use it to develop models and algorithms, which we then build into software products. This platform makes the entire process simpler, faster, and more streamlined.”

Mogens Mikkelsen

Enterprise Architect, SEGES Innovation

“As more of our groups rely on the Azure Machine Learning solution, our finance experts can focus more on higher-level tasks and spend less time on manual data collection and input.”

Jeff Neilson

Data Science Manager, 3M

A welder working

i

“With Azure Machine Learning, we can show the patient a risk score that is highly tailored to their individual circumstances. …Ultimately, we aim to reduce risk, reduce uncertainty, and improve surgical outcomes.”

Professor Mike Reed

Clinical Director, Trauma & Orthopedics, Northumbria Healthcare NHS Foundation Trust

A medical professional speaking with a patient

1

“The ability to scale compute resources up and down is critical for innovation speed and cost efficiency…. Azure Machine Learning and its built-in machine learning operations capabilities make agility and cost-efficiency simple.”

Kate Puech

Director of AI Engineering, Axon

.

“Using the automated machine learning features of Azure Machine Learning for machine learning model creation enabled us to realize an environment in which we can create and experiment with various models from multiple perspectives.”

Keiichi Sawada

Corporate Transformation Division, Seven Bank

A Seven Bank location
Back to tabs

IDC MarketScape: MLOps 2022 Vendor Assessment

Learn how enterprise organizations across industries are using MLOps to overcome the challenges of implementing AI and machine learning technologies.

Engineering MLOps white paper

Discover a systematic approach to building, deploying, and monitoring machine learning solutions with MLOps. Rapidly build, test, and manage production-ready machine learning lifecycles at scale.

Forrester Total Economic Impact study

The Forrester Consulting Total Economic ImpactTM study, commissioned by Microsoft, examines the potential return on investment enterprises may realize with Azure Machine Learning.

Machine Learning solutions white paper

Learn how to build more secure, scalable, and equitable machine learning solutions.

Responsible AI white paper

Learn how to build more secure, scalable, and equitable machine learning solutions.

MLOps white paper

Accelerate the process of building, training, and deploying models at scale.

Azure Arc–enabled machine learning white paper

Learn how to build, train, and deploy models in any infrastructure.

Frequently asked questions about Azure Machine Learning

  • The service is generally available in several countries/regions, with more on the way.

  • The SLA for Azure Machine Learning is 99.9 percent uptime.

  • Azure Machine Learning studio is the top-level resource for Machine Learning. This capability provides a centralized place for data scientists and developers to work with all the artifacts for building, training, and deploying machine learning models.

Ready when you are—let's set up your Azure free account

Try Azure Machine Learning free
OSZAR »