Featured Stories

Filter By Categories

Azure Machine Learning Lands in General Availability

By Lars Klint  |  December 07, 2018  |   azure Machine Learning New Products   |  

The field of machine learning (ML) is one that has grown exponentially in the last decade, both in capability as well as in number of engineers focusing on it. The problem for a long time has been that ML has been difficult to crack and get into. Businesses are pretty sure they want to use ML and AI in their processes, but they don’t know where to start. How do they drive AI and ML in the software they are building for tomorrow that includes automation and intelligent prediction?

This week Microsoft outlined the next phase of their ML platform on their Azure cloud platform at the Connect(); event and it is looking mighty promising. This update was geared towards making it easier to use the models and output from Azure ML Services, and connect it into your app that you use.

azure-ml-blog

As a developer this is hugely exciting. It means I now have roughly three steps to get ML models and automation into my app.

  1. Create a workspace: You install the Azure ML SDK and create your workspace in the cloud, where you can store your resources, models, deployments and results.

  2. Training the model: Let Azure help you train your model and identify suitable algorithms and parameters much faster.

  3. Deployment: Deploy the model to generate predictions. Use the output in your applications with the ML.NET toolkit for .NET developers.

Of course, you can still create your own models, which is predominantly done in Python using Jupyter notebooks. On top of this there is support for native APIs to connect to storage and compute resources that you will most likely need.

But it doesn’t end there. Part of the announcement was support for Kubernetes and container services, which is the serverless approach to ML.

Personally, I am very excited to see this big investment in the Azure ML platform, and it seems Microsoft is continuing the pace. The ML workflow now better resembles what large companies want out of the technology and I am sure we will see and hear a lot more about the platform.

To learn more about Azure Machine Learning services, read the official documentation and get started today with the 5-minute quickstarts.

Popular Tags

Enterprise Career Certification News Partner Story New Products Press Releases Corporate Interviews Cloud Security Google Cloud Platform Serverless AWS AWS Summit GCP Machine Learning azure tradeshows

Search the ACG blog