MICROSOFT ANNOUNCES NEW AZURE AI PLATFORM INNOVATIONS

Microsoft recently shared some new innovations for developers within its Azure AI Platform. The updates are meant to help businesses and organizations augment their own ingenuity and digital transformation efforts through new features related to search and machine learning capabilities. Here’s a quick rundown of some of the changes.

Cognitive Search
Cognitive Search is a new capability within the Azure Portal that uses AI to understand content and then integrate that information into Azure Search. It supports built-in file readers like PDFs and Office documents, cognitive skills for OCR like handwritten or print text, entity recognition, key phrase extraction, language detection, image analysis, and even facial recognition. So essentially, Search can pull data from a variety of Azure sources and use real knowledge and cognitive skills to organize and store the data in an index that’s optimized for intuitive exploration.

Azure ML SDK for Python
Within Azure Machine Learning, this SDK offers a single control plane API that allows developers or data scientists to execute the key AML workflows of Provisioning Compute, Model Training, Model Deployment and Scoring directly within Python.

Azure ML Packages
Microsoft is also now offering Azure ML Packages that represent a rich set of Python pip-installable extensions to Azure ML. These are meant to accelerate deep learning scenarios and make building highly efficient models more streamlined.

ML.NET
ML.NET is another new innovation in the Azure AI Platform that is now available in preview. The cross-platform open source framework is meant for .NET developers and provides an enterprise-grade software library of the latest and most talked about innovations in Machine Learning from Microsoft platforms and teams, including Bing, Office, and Windows.

Azure ML with Project Brainwave
Project Brainwave, which is also now available in preview, is getting integrated with Azure ML as well. Brainwave’s architecture is built for deep neural net processing, so it uses hardware acceleration to enable fast AI.

Featured image: Freerange Stock