Microsoft Build 2023: The Biggest Azure News

Azure AI Content Safety, Azure Cognitive Service for Language and Microsoft Fabric were included in Microsoft’s biggest Azure news at Build 2023.

A preview for Azure AI Content Safety. The general availability of ready-to-use document and conversation summarization for Azure Cognitive Service for Language. And the reveal of Microsoft Fabric.

These are among the biggest Azure announcements during the vendor’s Build 2023 conference. Microsoft is holding its annual developer-focused conference online Tuesday and Wednesday and in Seattle until Thursday.

The investments in generative artificial intelligence come as the Redmond, Wash.-based vendor positions the technology as the next popular user interface, with natural language interactions by users as revolutionary as graphical user interfaces.

“It’s not an autopilot approach, but rather a design ethos that we call ‘copilot’ to tackle the growing volume of digital debt that is taking attention away from innovation and sapping our productivity,” Rajesh Jha, Microsoft executive vice president of experiences and devices, said in a blog post.

[RELATED: Microsoft 365 Copilot Brings Generative AI To Teams, Word And More]

The generative AI race has mostly focused on Microsoft and Google’s efforts, with vendors including Salesforce, IBM and Palo Alto Networks all incorporating the technology in various offerings to speed up data analysis and boost security, among other use cases.

Solution providers have also experimented with the new technology to find more ways to optimize their business and find new ways to bring in revenue.

In an example of generative AI’s popularity, Azure OpenAI Service is still in preview but has been used by more than 4,500 companies, according to Microsoft.

Microsoft Executive Chairman and CEO Satya Nadella said on the vendor’s quarterly earnings call in April that Microsoft is already seeing new customers and new money-making opportunities from generative AI.

“Some of the work we’ve done in AI even in the last couple of quarters, we are now seeing conversations we never had,” he told analysts on the call. “Whether it’s coming through even just OpenAI’s APIs, right—if you think about the consumer tech companies … they have gone to OpenAI and are using their API. These are not customers of Azure at all.”

He continued: “Even Azure OpenAI API customers are all new. And the workload conversations, whether it’s B2C conversations in financial services or drug discovery … these are all new workloads that we really were not in the game in the past, whereas we now are.”

Other Azure news from Microsoft at Build 2023 includes:

*Updates to Azure development operations (DevOps) tools

*Advancements in Azure Database offerings

*Upgrades to Azure Infrastructure

Read on for more Azure news from Microsoft Build 2023.

Microsoft launched a preview for Azure AI Content Safety, a service meant to help developers make AI models that detect inappropriate content across images and text.

Billing for Azure AI Content Safety use starts June 1. The price is $1.50 per 1,000 images and 75 cents per 1,000 text records, according to Microsoft.

The models assign severity scores to the flagged content for human moderators to view, according to the vendor.

AI Content Safety is integrated into Azure OpenAI Service, which is in preview. Users can apply the service to online communities, gaming platforms and other non-AI systems. Users can fine-tune filters for context.

Along with AI Content Safety, Microsoft will release new media provenance capabilities for Microsoft Designer and Bing Image Creator. This will allow users to verify if AI generated an image or video through cryptography that marks and signs content with metadata about its origin, according to the vendor.

The media provenance will follow Coalition for Content Provenance and Authenticity (C2PA) standards.

During Build 2023, Microsoft unveiled Fabric, an end-to-end, unified analytics, Software-as-a-Service platform that brings together Data Factory, Synapse and Power BI.

Microsoft promises users that they don’t need to work with multiple vendors to get a complete analytics platform. Once Copilot is available for Fabric, users can leverage conversational language for data flows, data pipelines, code generation, machine learning (ML) models and visualizations.

Public previews available for Microsoft Fabric include Data Factory, which provides more than 150 connectors to cloud and on-premises data sources. Data Factory also promises drag-and-drop experiences for data transformation and the ability to orchestrate data pipelines, according to Microsoft.

Public previews are also available for Synapse Data Engineering, which users can leverage to author Spark experiences and instant start with live-pools, according to the vendor. Synapse Data Science promises users the ability to build AI models and train, deploy and manage ML models.

More public previews in Microsoft Fabric include Synapse Data Warehousing—a converged lake house and data warehouse experience—and Synapse Real-Time Analytics, which allows developers access to data streaming in from connected devices, telemetry, logs and other resources. Synapse Real-Time Analytics works on massive volumes of semi-structured data with high performance and low latency, according to Microsoft.

And in private preview is the data activator, which promises a no-code experience for real-time data detection and monitoring, according to Microsoft. Data activator can trigger notifications and actions when it finds specified patterns.

In an effort to eliminate data silos, Microsoft Fabric has a OneDrive-like multi-cloud data lake called OneLake available automatically to every tenant, with all workloads wired into OneLake as all M365 apps are wired into OneDrive. OneLake data is automatically indexed for discovery, sharing, governance and compliance.

One Lake is built on and compatible with Azure Data Lake Storage Gen2 (ADLSg2). It supports structured data of any format and unstructured data, according to Microsoft.

Its shortcuts capability allows for simplified data sharing among users and apps without moving and duplicating information.

Shortcuts can virtualize storage in ADLSg2 and Amazon Web Services (AWS) Simple Storage Service (S3). Compatibility with Google Storage is coming soon, according to Microsoft.

Microsoft will soon release in general availability (GA) ready-to-use document and conversation summarization for Azure Cognitive Service for Language.

The feature promises users the ability to summarize with Azure Service Level Agreements, data security and other Azure-backed enterprise-grade readiness, according to the vendor.

Now in preview for Azure Cognitive Service for Language is the ability to customize summarization.

An update for Azure Cognitive Search coming soon to preview will allow search by vector for improved storage, indexing and retrieval by concept of large language models apps.

Users can leverage keywords, text, images, audio, video and graphs for retrieving apps, according to Microsoft.

Microsoft made Document Translation in Language Studio GA, allowing developers to batch-translate documents.

Meanwhile, new previews include a Power Automate Translator connector and Text Translator software development kit (SDK) for .Net, Python, Java and JS.

The Power Automate Translator connector promises developers access to automation for translating text and documents stored across multiple clouds, on-premises and on local storage, according to Microsoft. More than 125 languages are available.

In private preview is an ACS Translator plugin for Azure OpenAI Service, according to Microsoft.

Microsoft plans to launch a preview of prompt flow for prompting, evaluating and tuning large language models.

The workflows can connect to various language models and data sources and assess workflow quality, according to Microsoft.

Another preview promises native capabilities to fine-tune and deploy foundation models from multiple open-source repositories using Azure ML components and pipelines. The feature will eventually support Azure OpenAI Service models.

A preview for a responsible AI dashboard that supports text and image data is available. Users can evaluate large models built with unstructured data during the model building, training and evaluation stage, finding model errors, fairness issues and model explanations before deployment, according to Microsoft. A preview for model monitoring promises performance tracking in production.

Microsoft has launched previews for a managed feature store, a Microsoft Purview connector, managed network isolation and DataRobot 9.0 support, according to the vendor.

Microsoft made Azure ML registries GA, which aim to help users promote, share and discover models, pipelines, environments and other ML artifacts across multiple workspaces in an organization.

Also GA is Azure Container for PyTorch. This feature aims to give users curated environments with PyTorch 2.0 capabilities, plus DeepSpeed, ONNX Runtime and other optimization software designed for large model training and inference, according to Microsoft.

Microsoft has a preview for a volume snapshot feature in Azure Container Storage, according to the vendor. This feature allows point-in-time capturing of states of persistent volumes.

An upcoming preview for Microsoft Cost Management should leverage AI for analysis and management of cloud costs and bills. Users can ask prompts to summarize invoices, break down costs, help reduce costs and explain bill fluctuations, according to the vendor.

In confidential computing, Microsoft has previews available for new confidential virtual machines (VMs) based on Intel Trust Domain Extensions and new AMD-based confidential VM options for Azure services.

Confidential containers on Azure Kubernetes Service (AKS) are in preview while ones on Azure Container Instances are now GA, according to Microsoft.

Red Hat Enterprise Linux (RHEL) support for Azure AMD-based confidential VMs became GA.

Microsoft made Azure Deployment Environments GA for self-deploying on demand. ADE comes with a portal for creating and managing environments.

In July, Microsoft Dev Box becomes GA. Developers will gain the option to view, manage and spin up dev boxes from a specialized portal. Certain users can use configuration-as-code files to customize boxes without administrative support. And users gain new starter images.

A preview for GitHub Advanced Security for Azure DevOps aims to give users natively embedded, automated security checks for code, secrets and supply chain, according to Microsoft.

Amd an Azure API Center is in preview while a self-hosted gateway support for Azure Active Directory (Azure AD) tokens is GA, among a host of other developer and DevOps updates.

New GA advancements to Azure Cosmos DB include burst capacity for using database or container idle throughput capacity for traffic spikes.

Hierarchical partition keys, also GA, increase the number of usable partition keys from one to three for improved data distribution and scale.

Microsoft launched a preview of Materialized Views for Azure Cosmos DB for NoSQL, which promises the ability to create and maintain secondary data views in containers used to serve queries. Those queries would be too expensive to serve with an existing container, according to the vendor.

A preview for Azure SQL Database Hyperscale elastic pools aims to help developers build and manage new apps in the cloud, according to Microsoft. Each pool holds up to 25 databases with up to 100 terabytes of storage.

LinkedIn shelved planned move to Microsoft Azure, opting to keep physical data centers

Microsoft CEO Satya Nadella arrives at federal court in Washington on Oct. 2, 2023.

Nathan Howard | Bloomberg | Getty Images

LinkedIn has set aside an effort to relocate its data center technology out of its physical facilities and into Microsoft's Azure cloud, according to people familiar with the matter.

The decision not to proceed with the project, code-named "Blueshift," marks a major reversal for LinkedIn, which announced its plan to move to Azure in 2019, three years after Microsoft acquired the company for $27 billion. LinkedIn had been using Azure for specific tasks.

The U-turn represents a setback for Microsoft, which is chasing Amazon Web Services in the lucrative cloud infrastructure market and has been counting on cloud technology and services to fuel much of its growth. Microsoft CEO Satya Nadella ran the cloud business before elevation to his current job in 2014.

Mohak Shroff, LinkedIn's vice president of engineering, wrote in a 2019 blog post announcing Blueshift that "moving to Azure will give us access to a wide array of hardware and software innovations, and unprecedented global scale."

Staffers started to learn of the decision not to follow through with the Azure migration last year, said the sources, who asked not to be named because of confidentiality. Executives stressed that the project was being put on hold, rather than getting canceled altogether, they said.

In a memo to research and development employees in June 2022, LinkedIn Chief Technology Officer Raghu Hiremagalur said LinkedIn would continue to use some Azure services and will "focus our efforts on scaling and innovating our on-prem infrastructure." A different internal document, viewed by CNBC, says LinkedIn and Microsoft together agreed to hold off on trying to get LinkedIn's website running on Azure.

"With the incredible demand Azure is seeing and the growth of our platform, we've decided to pause our planned migration of LinkedIn to allocate resources to external Azure customers," Hiremagalur wrote in his memo.  

A LinkedIn spokesperson confirmed that the Microsoft subsidiary changed direction on Blueshift and said LinkedIn continues to use Azure.

"We are using both Azure to complement our infrastructure needs and further investing in our data centers," the spokesperson said in an email. "This includes our running 100 employee-facing applications on Azure, leveraging Azure FrontDoor and ongoing work to consolidate our datacenter locations that are currently spread across multiple buildings under a single roof. Azure has been crucial to support and scale collaboration and productivity for our teams and to deliver value to our members." 

Azure Front Door is a content delivery network that keeps information stored in a variety of places around the world so it can quickly be sent to devices.

Issues with the planned migration arose from LinkedIn attempting to use its own software tools instead of those readily available on Azure, one of the people said. LinkedIn is in the process of constructing an additional data center to handle its computing needs, the person said.

Under the leadership of Nadella, Microsoft has moved some of its acquired assets to Azure, including GitHub and Minecraft developer Mojang.

More recently, Azure has gained attention because of Microsoft's investment in OpenAI, which uses Azure infrastructure for running the large language models powering ChatGPT and other products. Nadella told Wired that he first saw the GPT-4 LLM from OpenAI in the summer of 2022, a few months before OpenAI released the ChatGPT chatbot.

Microsoft said in October that third-quarter revenue from Azure and other cloud services grew 29%, while LinkedIn revenue was up 8%. LinkedIn said in November that it had reached 1 billion members.

WATCH: January's hiring rate is expected to be down from previous years, says LinkedIn's Karin Kimbrough

watch now