The Intelligent Data Platform oversees all of the company’s existing tools for data: SQL Server, Azure SQL, Aure Cosmos DB, Azure Synapse Analytics, Power BI, Purview and Azure AI.
Most of these services received input during the Build. SQL Server 2022 for example, still in preview, can now be used with Azure Synapse Link, Azure Synapse Analytics and Microsoft Purview, always with this idea of having more returns on the stored data.
According to Microsoft, these are strong requests from testers who have integrated the Early Adoption Program of SQL Server 2022. This is also the case for the link function of Azure SQL Managed Instance, with the idea of taking over the benefits of a PaaS (platform as a service) and apply them to disaster recovery.
Ledger is another example. It allows you to cryptographically link data from a relational database to a blockchain to make it verifiable. In particular, Ledger must simplify scenarios that include multiple parties in the processes, each of which can verify the integrity of the central data. It can also be used in audits, the tool proving cryptographic proofs of integrity.
Many other improvements can be found in almost every service. Azure SQL Database, for example, now allows local development through an emulator and a containerized runtime. Extensions for Visual Studio Code and Azure Data Studio are also available.
Flexible Server is a new offering from Azure Database for MySQL to accelerate provisioning, connecting, and development through tools already integrated with GitHub, Terraform, Azure Kubernetes Services, and Web Apps. The objective, of course, is to allow MySQL connoisseurs to continue to use their knowledge… in the Microsoft cloud. It’s actually general availability, with a new segment, Business Critical, distinguished by its low latency, high availability, and improved load balancing.
Power Platform strengthens its “low code” offensive
Power is the other big platform to have benefited from a series of announcements. It has been pushed by Microsoft for several years as the solution for rapidly developing applications that can be used in business, in interaction of course with the publisher’s services. It is in particular its “low code” aspect that is highlighted.
Microsoft has largely refocused on this. For example, Apps Portal is replaced by Power Pages, for the simplified creation of websites designed for desktop and mobile. Pages offers a design studio, templates, a series of tutorials, all integrating with Visual Studio, GitHub, and Azure DevOps. Security and governance specific tools have been added.
Azure Cognitive Services are leveraged to be able to retrieve a simple drawing made on a white sheet. It is sent to Express Design to start designing the page organization taking into account what was done in pencil. Manipulation is also possible from a PDF, a PowerPoint or even from Figma. Express Design will then suggest an interface and data model based on what was detected. Obviously, the effectiveness of the process depends on the effectiveness of recognition.
Still with this simplification in mind, Virtual Agents will be merged with Azure’s Bot Framework Composer. Microsoft assures that users of low-code and pro-code approaches will be able to collaborate effectively.
On the Power BI side, Datamart is entering, in preview for the moment. As the name suggests, it is a self-service type capability for discovering useful and actionable information in datasets, dependent or not. People with the appropriate access rights will therefore be able to create their own relational databases for analysis purposes, always from a low-code perspective. Microsoft no longer hides its desire to create a turnkey data storage solution directly in Power BI.
As for Power Automate – dedicated to the rationalization and automation of repetitive tasks – you can now delegate the monitoring of RPA (robotic process automation) tools to virtual machines in Azure, via the starter kit for Azure Virtual Desktop.
Rich preview and integration with Power Apps in Teams
Teams is still at the center of Microsoft’s business strategy. Since its release, not a publisher conference has gone by without one or more announcements about it, and Build 2022 was no exception.
Developers will soon be able to offer the deployment of rich links in Teams based on Adaptive Cardsn, whether to sites or web applications. Manifests will need to be modified accordingly. The preview is scheduled for this summer.
One of the most important changes – and which could have been in the previous chapter – concerns the integration of Teams with the Power platform. Translation, elements of Teams such as Chat, Meetings, Tasks, Files, or even Approvals become integrable into Power Apps, as well as the collaborative functions of Microsoft 365. The Graph API for Teams can also be used. Again, the preview is due to arrive this summer.
And as a reminder of how serious the company is about Teams, Microsoft introduced Live Share. The function will allow third-party publishers to offer collaboration services passing through Teams. For example, a demonstration was made on Hexagon’s Live Share module, which was used to modify and annotate a 3D model collaboratively.
Finally, the Teams Toolkit for Visual Studio Code and TeamsFX CLI are available in final versions. The Teams App Store has also been updated to make item discovery more efficient. The recommendations are intended to be more intelligent and there is a section dedicated to applications put forward by Microsoft.
Azure is still shooting in all directions
What about Azure then? The announcements were once again numerous. The company’s cloud offering continues to expand, resulting in years of double-digit quarterly growth for the cloud division. So no question of slacking off.
We start with two new features for Azure AI. First a preview of Azure OpenAI Service, which enables the use of OpenAI language models – including those based on GPT-3 – in a number of scenarios, including understanding, code generation and support to writing. On the other hand, the update of the Cognitive Service for Language to bring it the support of the personalized names of entities.
We stay in artificial intelligence with the Responsible AI dashboard for Azure Machine Learnig, oriented – as its name suggests – towards the responsible construction of AI solutions. The service is equipped with forward-looking dashboards to visualize the impact of AI in a given solution, through a series of statistics. In the example given by the firm, the English National Health Service uses it to identify patients at increased risk during operations.
Several new services are appearing in preview. For example, Form Recognizer is dedicated to documents and allows the detection of a greater number of elements. Metrics Advisor allows for its part to identify anomalies and define sensitivity thresholds, among other things.
Some products that were in preview are on the contrary in the final version and are therefore officially available. This is the case of Azure Container Apps, built on Kubernetes and allowing the execution of code from any type of Linux-based container. Same thing for the Azure Communication Services Mobile UI Library (yes, Microsoft keeps its “knack” for the names of its products).
We should also mention the renaming of Azure Spring Cloud to Azure Spring Apps, the arrival of Draft 2 for the faster publication of deployable applications in Azure Kubernetes Service (AKS), Draft Azure CLI, or even a Kubernetes-based Event Driven extension Autoscaling (KEDA) for AKS.
Azure App Service receives several improvements. First, the arrival in final version of the Landing Zone Accelerator, which should simplify the migration of on-site installations to the cloud, via a combination of documentation and automation.
There are also two new features in preview. First, support for Google Remote Procedure Call (gRPC), which uses HTTPS/2 to streamline message traffic between clients and servers. Next, batch migration capabilities that will arrive in the coming weeks. They will identify ASP.NET web applications, categorize those ready for migration, suggest SKUs (Stock Keeping Units) for them, and provide guides.
Finally, here is a list of miscellaneous improvements for Azure:
- The final release of agentless discovery and clustering of Hyper-V machines and physical servers, to ensure all required components are identified and included before migrating to Azure
- Preview improvements for estimation by Azure SQL, including recommendations for SQL Server in Azure Virtual Machines, as well as support for Hyper-V machines and physical stacks
- The arrival in preview of a pause/resume capability for migrations, for example to schedule operations during off-peak hours
- The final availability of a service for discovering, estimating and modernizing ASP.NET web applications before moving them to Azure Application Service
- The ability, in the final release, for customers to take advantage of DCsv3 virtual machines with Intel Software Guard Extensions (SGX)
- Azure Stack HCI in final release
- A partnership with F5 to develop NGINX for Azure, in preview
- Partnering with Dynatrace to launch Dynatrace for Azure, which will proactively detect and resolve issues in critical applications