Tens of thousands of customers, such as Pepsi, Amway, Airbus, BBC, and Progressive Insurance are using Azure AI to deliver immersive customer experiences, identify new business opportunities, and drive impact. As more organizations use AI in their production workloads, they have an increased need for secure, scalable, and responsible AI development. Today we’re announcing new Azure AI capabilities that will empower organizations to use AI with even more confidence.

  • Train machine learning (ML) models on-premises and in multi-cloud environments with Azure Arc support for Azure Machine Learning.
  • Deliver cutting-edge search experience in your applications with the semantic search capability in Azure Cognitive Search.
  • Extract information from invoices and identification documents with new capabilities and expanded language support in Form Recognizer.

Build machine learning models anywhere with Azure Machine Learning

Customers can now use any Kubernetes cluster for model training with Azure Machine Learning—including on-premises, in multi-cloud environments, and at the edge— with the preview of Azure Arc enabled machine learning. This means organizations can extend machine learning to run close to where the data lives. They can lower cost by leveraging their existing Kubernetes infrastructure and increase operational efficiency. In addition, customers can manage all the resources through a single pane with the management, consistency, and reliability provided by Azure Arc. To learn more about all our Azure Arc announcements, read this blog.

We recently released capabilities for data professionals working in Azure Synapse to seamlessly collaborate with data scientists using Azure Machine Learning. Building on this collaboration, we are now announcing support for large scale, interactive data preparation, in Azure Machine Learning, backed by Azure Synapse Spark, in preview. Data Scientists can now prepare data and do feature engineering on big data directly from Azure Machine Learning notebooks, saving time, and optimizing workflows.

Deliver a highly relevant search experience with Azure Cognitive Search

Most organizations depend on in-app search as a critical component of their customer experience. Search services in the market today are syntax or keyword-based which produce results based on frequency of the term appearing in documents and not users’ intent. The new semantic search capability in Azure Cognitive Search enables organizations to provide a highly relevant search experience in their apps, including summary answers so users can get to what they need faster. Semantic search is based on years of AI research by Microsoft Research and uses some of the most advanced natural language models and techniques that are already being used in Bing search. With this capability, Azure Cognitive Search is the only cloud search service that enables organizations to deliver web-search type experience in their apps. Together with the largest region coverage, extensive language support, and industry leading security capabilities, Azure Cognitive Search is built to address your most critical business needs.

Azure Cognitive Search Mechanics video

“Our search solution “Smart Records” developed in co-creation with OrangeNXT and powered by Azure Cognitive Search, empowers us to answer more than 90 percent of all sales related questions quickly, resulting in shorter time to quote and better tailored proposals for our customers.” —Chris van Ravenswaay, Global Solutions Manager, Howden Group Ltd.

Extract information with new capabilities and expanded language support in Form Recognizer

Starting today, customers can now automatically extract data from invoices, as well as passports, drivers’ licenses, and other common forms of identification with Form Recognizer support for invoices and identification documents (IDs). This means organizations will be able to minimize manual data entry while improving customer experiences such as online banking transactions and hotel registrations. Customers will also be able to process forms submitted in more languages, with Form Recognizer expanding support to a total of 73 languages (an addition of 64 languages).

At GEP, we believe that 80 percent of all procurement and supply chain operations can be automated. By combining our AI solution with Form Recognizer, we automated the processing of 4,000 invoices a day for a client, saving them tens of thousands of hours of manual effort, while improving accuracy, controls, and compliance on a global scale.” —Sarateudu Sethi, GEP’s Vice President of Artificial Intelligence

AI is driving transformation across industries and we are humbled by the role our technology is playing in making that happen. We continue to look forward to how organizations use these new technologies to deliver impact.

Get started


Azure. Invent with purpose.

Similar Posts