On the first day of its Connect () 2017 conference in New York City, Microsoft announced a large number of new tools for machine learning applications in IoT devices, Apple’s Core ML and Azure SQL. But the one thing that seems to have captured the attention of many is the beta version of an extension meant for the company’s Visual Studio 2017 IDE. Called ‘Visual Studio Tools for AI’, the extension lets developers and data scientists embed deep learning models into applications and, supports deep learning frameworks such as Microsoft’s Cognitive Toolkit (CNTK) and Google’s TensorFlow, among others. It is already  available for download from the Visual Studio Marketplace. Microsoft is also releasing a similar extension called ‘Visual Studio Code Tools for AI’ for its free and open source (FOSS) code editor, Visual Studio Code.

The Redmond giant had earlier unveiled the extensions at its Ignite IT Pro conference last September in Orlando. Apart from the two deep learning frameworks mentioned above, Visual Studio Tools for AI and VS Code Tolls for AI support Caffe2, Theano, Keras and more. The two extensions enhance Visual Studio’s contextual editing capabilities and, feature Azure Machine Learning integration that allows users to “transfer AI model training jobs to Microsoft’s cloud” when the application needs extra processing power. The two extensions are part of Microsoft’s ‘Open Mind Studio’ suite that was announced last year.

Apart from the two extensions which now promise to enable developers to use AI services from inside Visual Studio, Microsoft also announced ‘Azure IoT Edge’, which “allows developers to deploy and run intelligent services on edge devices”. It is supported by Azure Machine Learning, Azure Functions and Azure Stream Analytics and, allows developers to run AI applications on the edge using containers. Microsoft also announced the commercial availability of a cloud-based solution for automating application lifecycles called ‘Visual Studio App Center’ that not only helps build, test and deploy apps faster, but also monitor real-world usage with crash and analytics data.