Machine learning operations (MLOps)
Accelerate automation, collaboration, and reproducibility of machine learning workflows
Streamlined deployment and management of thousands of models across production environments, from on premises to the edge
Fully managed endpoints for batch and real-time predictions to deploy and score models faster
Repeatable pipelines to automate machine learning workflows for continuous integration/continuous delivery (CI/CD)
Continuously monitors model-performance metrics, detects data drift, and triggers retraining to improve model performance
Deliver innovation rapidly
MLOps—machine learning operations, or DevOps for machine learning—is the intersection of people, process, and platform for gaining business value from machine learning. It streamlines development and deployment via monitoring, validation, and governance of machine learning models.
Build machine learning workflows and models
Use datasets and rich model registries to track assets. Enable enhanced traceability with tracking for code, data, and metrics in run history. Build machine learning pipelines to design, deploy, and manage reproducible model workflows for consistent model delivery.
Easily deploy highly accurate models anywhere
Deploy rapidly with confidence. Use managed online endpoints to deploy models across powerful CPU and GPU machines without managing the underlying infrastructure. Package models quickly and ensure high quality at every step using model profiling and validation tools. Use controlled rollout to promote models into production.
Efficiently manage the entire machine learning lifecycle
Take advantage of built-in interoperability with Azure DevOps and GitHub Actions for seamlessly managing and automating workflows. Optimise model training and deployment pipelines, build for CI/CD to facilitate retraining, and easily fit machine learning into your existing release processes. Use advanced data-drift analysis to improve model performance over time.
Achieve governance across assets
Track model version history and lineage for auditability. Set compute quotas on resources and apply policies to ensure adherence to security, privacy, and compliance standards. Use the advanced capabilities to meet governance and control objectives and to promote model transparency and fairness.
Benefit from interoperability with MLflow
Build flexible and more secure end-to-end machine learning workflows using MLflow and Azure Machine Learning. Seamlessly scale your existing workloads from local execution to the intelligent cloud and edge. Store your MLflow experiments, run metrics, parameters and model artefacts in the centralised Azure Machine Learning workspace.
Accelerate collaborative MLOps across workspaces
Facilitate cross-workspace collaboration and MLOps with registries. Host machine learning assets in a central location, making them available to all workspaces in your organisation. Promote, share, and discover models, environments, components, and datasets across teams. Reuse pipelines and deploy models created by teams in other workspaces while keeping the lineage and traceability intact.
Resource center
See MLOps in action
Build ML pipelines to design, deploy and manage model workflows
Build ML pipelines to design, deploy and manage model workflows
Automating a machine learning pipeline
Sam is a data scientist working for an online fashion retailer. The company recently started using machine learning to provide timely and accurate recommendations, improving both customer satisfaction and sales.
After a catalogue update, Sam was asked to update the brand recommendation model – but nobody knows where the current model came from. He turns to Azure Machine Learning to build a reproducible and traceable workflow.
Sam is a data scientist at an online retailer, tasked with updating their recommendation model after a recent catalog update. Sam is looking for the model pipeline, but no one knows where the model came from.
Sam turns to Azure ML to build a traceable model workflow.
Use Azure Machine Learning with your existing tools
Sam uses the Azure Machine Learning SDK to move his Jupyter Notebook–based work to Azure Machine Learning, taking advantage of its scale and collaboration capabilities. Using his choice of framework, he versions his training code and datasets and links them to a traceable experiment.
Capture the metrics from every experiment run
Every time Sam executes a run, Azure Machine Learning tracks its metrics and logs it for review. This captures the inputs (such as datasets, code, and parameters) and outputs (such as logs, metrics, and models) for assessment and comparisons.
Review models and experiments in the workspace
While Sam continues to work in his notebook environment, his team uses the Azure Machine Learning studio to easily trace back from the final model in the model registry through each asset and step used to create it.
Create reusable pipelines for efficiency and consistency
Sam creates a reusable machine learning pipeline to deliver a model training workflow and trigger it automatically upon changes in source data. If needed, he can reuse the outcome of steps without changing the inputs.
Build reproducible machine learning pipelines
Sam has built a reproducible and traceable workflow using Azure Machine Learning. Each model that the team produces is now tied to an experiment run for review and iteration. With a reproducible pipeline, Sam lets Azure Machine Learning automate retraining while he focuses on the next big thing.
Azure Machine Learning helps you build enterprise-grade machine learning pipelines through reproducibility and traceability. This leads to more consistent model delivery with less variability and increased fault tolerance.
Learn more about machine learning pipelines Browse the GitHub repo for this demo
By leveraging Azure ML, Sam has built a reproducible and traceable workflow. Each model that the team produces is now tied to an experiment run, giving the team the ability to review and iterate. With a reproducible pipeline, Sam can let Azure ML automate retraining while he focuses on the next big thing.
Azure ML helps you build an enterprise-grade machine learning pipelines through reproducibility and traceability. This leads to more consistent model delivery with less variability and increased fault tolerance.
Learn more about machine learning pipelines Browse the GitHub repo for this demo
Deploy rapidly with confidence, using auto-scaling, managed, distributed inference clusters
Deploy rapidly with confidence, using auto-scaling, managed, distributed inference clusters
Deploying a web service endpoint for a model
Sam, a data scientist working for an online fashion retailer, has added Azure Machine Learning to his workflow to build better models. His automated pipeline just registered a final model with attached experimental results showing improved performance.
The web development team isn't sure how to embed a machine learning model in the web app and has asked for a web service instead. Sam turns to Azure Machine Learning for help adding a web service endpoint to his model.
Sam, a data scientist at an online retailer, used Azure ML to create and register a model. His web dev team has now asked him for a web endpoint to call his model for inferencing. Sam uses Azure ML to create this web service endpoint.
Fetch the latest model and profile it for production
Sam fetches the latest model from the registry and profiles it to see what kind of resources it might need in production.
Push the model to Azure Container Instances
With the profiling data in hand, Sam uses Azure Machine Learning to package the model into a container and deploy it to Azure Container Instances for real-time or batch inferencing. After deployment, he takes advantage of the automatic schema generation and monitoring features in Azure Machine Learning.
Deploy to scalable, diverse endpoints and implement A/B testing
With just a couple lines of code, Sam pushes the same model to a more scalable Azure Kubernetes Service (AKS) instance and exercises advanced A/B testing capabilities for new model validation. He can now deploy models wherever they’re needed, from the cloud to the edge, using compute options including CPU, GPU, and FPGA.
Enable the web dev team to use the deployed model
After the model is deployed to an AKS instance, the web dev team accesses the web service endpoint to call from their application for real-time or batch inferencing needs. Sam monitors the model endpoint for data drift and other metrics, which can be audited as needed.
Publish models wherever they're needed
Sam turned his open-source model into a production-ready web service for his development team. He kept his focus on model development rather than learning the many technologies needed for service hosting, including Kubernetes, Flask, and Swagger.
Azure Machine Learning helps you seamlessly transition your models into mature and fully featured web services, enabling them to be rapidly adopted in applications.
With Azure ML, Sam has taken his open source model and quickly turned it into a production-ready web service ready for his development team to use. As a data scientist, he has been able to keep his focus on model development, rather than having to upskill on the many technologies needed for service hosting (including Kubernetes, Flask and/or Swagger).
Azure ML helps data scientists seamlessly transition their models into mature and fully featured web services, enabling them to be rapidly adopted in applications.
Interoperate with Azure DevOps and GitHub Actions to automate machine learning workflows
Interoperate with Azure DevOps and GitHub Actions to automate machine learning workflows
Integrating with your DevOps workflow
Sam, a data scientist working for an online fashion retailer, used Azure Machine Learning to build and deploy his models. He recently released his model to a production instance of Azure Kubernetes Service (AKS), and the web team has integrated it into the web app with promising results.
Rebecca, who's on the operations team, wants to ensure that model build and deployment is part of the DevOps workflows. Sam isn't sure how to incorporate it, so he trains Rebecca on Azure Machine Learning and works with her to determine how it can satisfy both their requirements.
Sam, a data scientist, recently used Azure ML to build and deploy models. He just released a model to their production AKS instance. Rebecca wants to integrate it with their DevOps workflows. Sam points her to Azure ML.
Create a new release pipeline with Azure Pipelines
Rebecca creates a pipeline in Azure Pipelines to automate the deployment process. The first step of the pipeline accomplishes tasks like building the model with all dependencies and checking for errors. To automatically trigger this pipeline, Rebecca connects it to the Azure Machine Learning model registry, ensuring that it starts whenever a new model or a new version of the model is registered.
Publish the model for validation
Rebecca adds a task to publish the model to Azure Container Instances for validation before it goes to production. She uses the Azure CLI to interact with Azure Machine Learning. It gives her access to the same features inside her tools that Sam has in his notebook. This step mimics integration tests in their app's DevOps process.
Move the model to production upon validation
Rebecca adds a release pipeline task to deploy the model to production using AKS. Before this task is activated, Rebecca configures a gate requiring a lead or manager to assess and approve before continuing. This ensures that the pipeline doesn’t cause major disruptions if an issue wasn’t caught in the scripted validations.
Setup data drift detection to trigger retraining
As Rebecca sets up the automated pipeline to deploy the models, Sam has enabled data drift monitoring with Azure ML. By monitoring service usage, he can see if the inference requests in production start to deviate from the training dataset, indicating a need to re-train.
Monitor service for stability and performance
Using Application Insights, a feature of Azure Monitor, Rebecca monitors service operations, such as number of requests, load on the AKS cluster and response latency, while Sam monitors model performance to determine whether to make optimisations.
Create a streamlined model workflow
Rebecca incorporated Azure Machine Learning into her DevOps workflow, from release automation through to production monitoring. While she keeps an eye on service operations, Sam monitors and manages model performance and optimises as necessary.
Azure Machine Learning helps elevate data science in your organisation's DevOps process, leading to more reliable and robust machine learning projects.
Azure ML has seamlessly integrated into Rebecca’s existing DevOps workflow, from release automation through production monitoring, with native integrations into existing tools like Azure DevOps and Azure App Insights. And while Rebecca keeps an eye on service operations, Azure ML provides Sam with tools to monitor and manage model performance and optimize as necessary.
Azure ML helps data science become a first-class citizen in your organisation’s existing DevOps process, leading to more reliable and robust machine learning projects.
Improve governance and cost management across your machine learning projects
Improve governance and cost management across your machine learning projects
Enabling machine learning model governance
Phil, the IT manager at an online fashion retailer, is enthusiastic about the work the data science and operations teams have been doing with Azure Machine Learning.
Although Phil is responsible for the company's IT infrastructure, he isn't sure what's in production, who has access, or how much it'll cost. He decides to use Azure Machine Learning to help him deliver on his governance responsibilities.
Phil, the IT Manager at an online retailer, is excited about Sam and Rebecca's work. But Phil is responsible for IT infrastructure and doesn't know how to ensure governance around the ML project. He's looking at Azure ML for help.
Review models in production
Phil uses Azure Machine Learning to audit the production endpoints. He can navigate back to underlying models and experiments along with the outputs of the model explanations and fairness analysis. This ensures that he traces every step in the model creation process and enables model auditing for any bias.
Check the audit logs for detailed model history
Phil accesses the Azure Machine Learning audit logs. They complement the traceability found in the models and experiments with a detailed log, which tracks every action (date, time and user). These logs help Phil trace actions initiated within or by the service for organisation-wide accountability.
Secure the workspace using Azure identity and access management
Phil uses Azure identity and access management solutions to manage access to the Azure Machine Learning workspace. He uses both built-in roles and customised roles to limit the tasks users can perform and control access to resources.
Ensure accountability with Azure resources
Phil uses the quotas capability to configure resource limits and sharing. This helps him manage spending, control costs, and ensure equitable resource allocation across teams. It also encourages accountability by the respective project teams.
Achieve model governance and compliance
Phil can now see what's in production and ensure organisational compliance through auditing, role-based access control and quota management. He appreciates how Azure Machine Learning not only supports the data science and operations teams, but also delivers the enterprise-grade governance capabilities he needs.
Azure Machine Learning provides the organisational controls essential for making machine learning projects successful and more secure.
Phil is thankful that Azure ML not only supports the data science and operations teams, but also delivers the enterprise-grade governance capabilities he needs to stay on top of his responsibilities. He can see exactly what is in production, as well as ensure organisational compliance through auditing, role-based access control and quota management.
Azure ML provides the organisational controls essential for making machine learning projects successful and secure.
Comprehensive security and compliance, built in
-
Microsoft invests more than USD 1 billion annually on cybersecurity research and development.
-
We employ more than 3,500 security experts dedicated to data security and privacy.
-
Azure has more certifications than any other cloud provider. View the comprehensive list.
Get started with an Azure free account
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 40+ other services that are always free.
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.
After 12 months, you'll keep getting 40+ always-free services—and still pay only for what you use beyond your free monthly amounts.
See how customers are delivering value with MLOps
FedEx
Bikram Virk, Product Manager, AI and Machine Learning, FedEx"Customers expect timely and accurate information on their packages and a data-based delivery experience. We're helping FedEx stay on the leading edge with Azure Machine Learning, and we're building expertise for future projects."
BRF
Alexandre Biazin, Technology Executive Manager, BRF"We're scaling with automated machine learning in Azure and MLOps capabilities in Azure Machine Learning so that our 15 analysts can focus on more strategic tasks instead of the mechanics of merging spreadsheets and running analyses."
Nestle
Ignasi Paredes-Oliva, Lead Data Scientist, Nestlé Global Security Operations Centre"MLOps is at the core of our product. Because of its reproducible ML pipelines, ...registered models, and automatic model scoring, we're definitely detecting things that we missed before. Which, in terms of risk management, is really, really important."
PepsiCo
Michael Cleavinger, Senior Director of Shopper Insights, Data Science, and Advanced Analytics, PepsiCo"We've used the MLOps capabilities in Azure Machine Learning to simplify the whole machine learning process. That allows us to focus more on data science and let Azure Machine Learning take care of end-to-end operationalization."
Additional MLOps resources
Get started
Learning
End-to-end MLOps learning pathBlog
- Unifying MLOps at Microsoft
- MLOps maturity model concepts
- Azure Machine Learning excels in enterprise readiness
- The art of testing machine learning systems
- Testing the robustness of machine learning systems
- Testing the scalability of machine learning systems
- Testing the security of machine learning systems