As companies contention to physique oregon modernize intelligent apps that incorporated generative AI, galore are uncovering gaps successful their tech-stack modernization efforts arsenic they research and trial usage cases. 

Recently, Microsoft commissioned Forrester Consulting to measure modernization efforts astatine planetary organizations. The survey recovered that astir 90% of determination makers said they person a modern tech instauration and are acceptable for the future. The survey besides recovered there’s a cognition spread and galore aren’t arsenic acceptable arsenic they deliberation they are. Some inactive woody with inflexible, outdated, and brittle bequest bundle systems that stifle innovation and permission important country for improvement. 

Some of this comes from a unreality attack focused connected migration. There is simply a batch to summation simply successful migrating to the cloud, but modernization is overmuch much than assistance and shift—it’s astir what you tin bash erstwhile you get there. 

The survey besides recovered that 1 successful 5 companies surveyed are overcoming foundational barriers successful their exertion modernization travel to bring captious worth backmost to the business. These companies are doing a fewer communal things: implementing a wide exertion strategy built astir AI and training, integrating analytics and cardinal metrics to link with concern outcomes, and upgrading exertion stacks, supported by strategical partnerships with applicable suppliers. In adopting these practices, these companies tin determination rapidly to code interior gaps and flooded thorny challenges successful the mode of their modernization goals. 

Is your tech stack AI-ready? Are you looking for champion practices to assistance execute your modernization strategy? Want to larn much astir the Forrester study? Join america connected May 8, 2024 for our adjacent Microsoft Azure webinar: “Harness the Power of Intelligent Apps: Modernize with Azure.” In summation to proceeding from me, you’ll perceive from: 

• Amanda Silver connected “Powering Future Innovation with Azure Application Platform”
• Ali Powell connected “Vision to Value: Accelerating Modernization to Azure with Azure Customer Success”
• Cyril Belikoff connected “Empowering Modernization with Microsoft Programs, Offers, and Partner Ecosystem (Closing Keynote)”

And we’ll beryllium joined by Bill Martorelli, Principal Analyst astatine Forrester, who volition sermon the study. 

Next, present is the latest from Microsoft data, AI, and integer apps to assistance you with your modernization.

What’s caller successful data, AI, and integer applications 

Snowflake concern extends Azure options for your data 

One of the astir empowering things Microsoft does for customers is forge partnerships with companies who widen the worth of our platform. Snowflake is simply a large illustration of this successful action. I had the accidental to articulation Snowflake Chief Marketing Officer Denise Persson to speech astir however we enactment unneurotic to assistance customers marque the astir of their information to thrust AI transformation. Partnerships similar these are conscionable 1 much crushed wherefore Azure is the champion unreality for your data. 

More choice—Microsoft Azure AI exemplary catalog of unfastened and frontier models conscionable got bigger with additions from Snowflake, Microsoft Research, Meta, Databricks, and more 

The Azure AI exemplary catalog now offers much than 1,600 instauration models (large connection models (LLMs) and tiny connection models (SLMs)) enabling Azure customers to take the champion exemplary for their usage case. Our catalog continues to grow, with the summation of caller cutting-edge models, including Snowflake Arctic, Meta Llama 3, databricks/dbrx-base and databricks/dbrx-instruct, and Microsoft’s Phi-3 household of unfastened models, the astir susceptible and cost-effective tiny connection models available, outperforming models of the aforesaid size and adjacent size up crossed a assortment of language, coding, and mathematics benchmarks. The caller models not lone grow our catalog, but besides heighten our quality to conscionable divers endeavor demands.  

More standard astatine nary further cost: Increased vector capableness successful Azure AI search 

Azure AI Search present has importantly accrued retention capableness and vector scale size astatine nary further cost. To enactment accrued request for retrieval augmented procreation (RAG) and generative AI applications astatine scale, AI Search has drastically raised retention limits, giving customers much vectors per dollar without compromising precocious performance. With this change, customers tin execute much scalability astatine a little cost, spot AI Search to grip their ample RAG workloads, and use precocious hunt strategies to navigate analyzable information to innovate successful ways antecedently unimaginable. 

Microsoft announces caller concern AI innovations from the unreality to the mill floor 

We announced caller AI-driven solutions for manufacturing that purpose to unify information estates, heighten operational resilience, and empower front-line workers with AI copilots for businesslike information querying and contented resolution. The backstage preview includes Microsoft Fabric information solutions and the Microsoft Copilot for Azure AI template, promising to optimize operations and accelerate AI integration crossed planetary manufacturing sites. 

New infrastructure for the epoch of AI: Emerging exertion and trends successful 2024 

From unreality to edge, CPUs to GPUs, and application-specific integrated circuits (ASICs), the AI hardware and bundle scenery is expanding astatine an awesome rate. Our caller “Microsoft Azure: The State of AI Infrastructure” study helps you support up with the existent authorities of AI, its trends and challenges, and to larn astir champion practices for gathering and deploying scalable and businesslike AI systems. Read the inaugural blog successful our caller Infrastructure for the Era of AI bid for much and to download the report. 

Mirroring successful Microsoft Fabric for Azure SQL Database and Azure Cosmos DB present successful nationalist preview 

Azure SQL Database Mirroring successful Microsoft Fabric continuously replicates your SQL information into Delta tables successful Microsoft Fabric’s OneLake successful adjacent real-time. This caller diagnostic helps debar analyzable and time-consuming extract, transform, load (ETL) processes and enables faster clip to insights. Mirroring further reduces your wide full outgo of ownership with zero compute and retention costs to replicate, helping you rapidly and cost-effectively respond to changes successful your business. 

With Azure Cosmos DB Mirroring successful Microsoft Fabric, you tin seamlessly bring your Azure Cosmos DB information into OneLake successful Microsoft Fabric with no-ETL and adjacent real-time insights connected your operational data, allowing you to respond much rapidly to changes successful your concern situation and marketplace conditions. 

Microsoft Azure Database for PostgreSQL: New AI capabilities successful Azure AI hold and caller migration service 

The Azure AI hold successful nationalist preview present includes real-time predictions to invoke instrumentality learning models hosted connected Microsoft Azure Machine Learning online endpoints. This is highly adjuvant for gathering fraud detection successful banking, merchandise recommendations successful retail, diligent predictions successful healthcare, and more. Also caller is real-time substance translation utilizing Microsoft Azure AI Translator to construe substance from wrong SQL. This facilitates gathering intelligent multilingual applications connected Azure Database for PostgreSQL. It besides supports automatically detecting the connection of the input substance and filtering profanities.  

A new migration work present mostly available simplifies the process of moving PostgreSQL databases from anyplace to Azure. The work offers some offline and online migration options from an on-premises server, AWS RDS for PostgreSQL, Azure Virtual Machines (VMs), and Azure Database for PostgreSQL—Single Server. The work helps customers determination to Azure Database for PostgreSQL—Flexible Server with easiness and confidence. 

Microsoft Defender for Cloud present supports Azure Database for MySQL—Flexible Server 

Protect databases from threats without affecting show oregon availability, and little the hazard of information breaches, attacks, and unauthorized entree with information monitoring of anomalous oregon suspicious activities. Customers tin easy alteration Defender for Cloud from the Azure Portal to commencement getting information alerts and insights for Azure Database for MySQL—Flexible Server and person recommendations to mitigate perchance harmful threats. Learn much astir Defender for Cloud. 

Healthcare information solutions successful Microsoft Fabric successful nationalist preview 

The healthcare information solutions supply information models and translation activities that assistance customers make a multi modal warehouse. It enables customers to align with manufacture standards, specified arsenic Fast Healthcare Interoperability Resources (FHIRs) and Digital Imaging and Communications successful Medicine (DICOM), and supports compliance with regulations, specified arsenic the Health Insurance Portability and Accountability Act of 1996 (HIPAA) and General Data Protection Regulation (GDPR). Learn much astir healthcare information solutions.

Empowering AI globally with investments successful Japan and United Arab Emirates, a caller AI Hub successful London, and a planetary concern with Cognizant 

Microsoft volition put USD2.9 billion implicit the adjacent 2 years to summation our hyperscale unreality computing and AI infrastructure successful Japan. We volition besides grow our integer skilling programs, unfastened our archetypal Microsoft Research Asia laboratory successful Japan, and deepen our cybersecurity collaboration with the Government of Japan. 

We are investing USD1.5 cardinal successful Abu Dhabi’s G42 to accelerate AI improvement and planetary expansion, and to fortify our collaboration connected bringing the latest Microsoft AI technologies and skilling initiatives to the United Arab Emirates and different countries astir the world. 

Microsoft AI Chief Executive Officer Mustafa Suleyman announced Microsoft AI is opening a caller AI hub successful the bosom of London. Microsoft AI London volition thrust pioneering enactment to beforehand state-of-the-art connection models and their supporting infrastructure, and to make world-class tooling for instauration models, collaborating intimately with our AI teams crossed Microsoft and with our partners, including OpenAI. 

Cognizant and Microsoft are expanding our partnership to integrate Microsoft’s generative AI and Microsoft Copilot into Cognizant’s integer translation services. With a important concern successful generative AI, the concern is acceptable to thrust AI adoption and make industry-specific solutions, portion upholding ethical AI standards. The inaugural besides includes extended grooming for Cognizant developers and the deployment of Microsoft Copilot for Microsoft 365 to a immense idiosyncratic base.  

Learn thing new 

Microsoft Learn helps code the AI skills gap 

With much than 6 cardinal radical globally engaged with AI learning, Microsoft Learn is your AI skill-building spouse successful addressing the AI skills gap. We designed a elemental model to assistance you illustration your ain people for gathering the indispensable AI skills to recognize the worth of the Microsoft platform. Learn much with the “Accelerate AI translation with accomplishment building” presumption paper. For more, spot the blog from Kim Akers, Corporate Vice President, Enablement and Operations for Microsoft Customer and Partner Solutions, connected however to alteration your concern with AI accomplishment gathering connected Microsoft Learn. 

How to customize your generative AI model 

There are 3 cardinal techniques to customizing a LLM: punctual engineering, RAG, and fine-tuning. In this blog, you volition larn however to:  

• Optimize prompts to guarantee a exemplary produces much close responses.  
• Retrieve accusation from outer sources for amended accuracy. 
• Teach a exemplary caller skills and tasks. 
• Teach a exemplary caller accusation from a customized dataset. 

LLMOps Maturity Model Assessment and framework, positive caller instructional video series 

LLMOps describes the operational practices and strategies for managing LLMs successful production. Not definite what your organization’s LLMOps maturity level is? Take a 10-minute assessment to find out. Then get guidance connected however to advance your capabilities successful LLMOps based connected your organization’s maturity level.  

There are a batch of intricacies progressive successful gathering systems that usage LLMs: selecting conscionable the close models, orchestrating LLM flows, and monitoring them utilizing liable AI toolchains. LLMOps is much than exertion oregon merchandise adoption. It’s a confluence of the radical engaged, the processes used, and the products implemented. In a new instructional video series, we volition research the conception of LLMOps successful depth, alongside introducing the latest Azure AI tools designed to facilitate the adoption of robust LLMOps practices for enterprises. 

New blog: Improving Azure Functions acold starts 

Our caller blog, “How to Conquer Cold Starts for Better Performance” is present unrecorded connected The New Stack, the level for the latest quality and resources for unreality autochthonal technologies. Skill-up for escaped connected leveraging AI with cloud-native and serverless technologies connected Azure.

Insights Tomorrow: A podcast for information enthusiasts 

Insights Tomorrow features in-depth conversations with information leaders and experts astir the revolutionary journeys they’re taking successful the satellite of data, analytics, and governance. Join big Patrick LeBlanc to research however unlocking worth from information is trailblazing the mode to integer transformation. 

Customers innovating with generative AI  

MultiChoice boosts idiosyncratic restitution with Azure Machine Learning and Microsoft Azure Databricks 

Have you ever wished for a viewing acquisition that knows your sensation amended than you do? MultiChoice, Africa’s premier amusement destination, has made that a world with Azure Machine Learning. Users bask highly personalized recommendations that cater to their tastes, making each presumption count. As MultiChoice is making the viewing acquisition better, smarter, and much enjoyable, it has seen a singular uptake successful idiosyncratic engagement and satisfaction. 

Nexi uses Microsoft Azure App Service to modernize its FinTech apps to supply a personalized outgo solution to customers 

Nexi is simply a European PayTech institution that helps its customers wage and judge integer payments done a implicit and personalized scope of simple, intuitive, and unafraid solutions. To simplify transactions and empower radical and businesses to bask person relationships and prosper together, Nexi needed a unified solution with much flexibility to let its developers to walk overmuch little clip connected infrastructure activities and supply much worth connected delivering products. Working with Microsoft and cVation, Nexi moved guardant to implement app modernization and transformation. 

Coles accelerates from monthly to play exertion deployments with Azure 

Coles, which operates much than 800 stores successful Australia, needed to modernize its exertion to conscionable increasing lawsuit request for an omnichannel acquisition that served them arsenic good successful store and online. In the highly competitory market business, that means rapidly evolving successful effect to lawsuit input. Coles gained astonishing efficiency with Azure DevOps, shifting from monthly to play deployments, reducing physique times, and enabling accelerated deployment of changes to production. The lawsuit acquisition is further optimized with Azure Cosmos DB, which provides an aggregated presumption crossed channels to alteration further lawsuit insights, allowing Coles to supply an adjacent much customized experience. 

Opportunities to connect 

Build multimodal apps with Microsoft and triumph astatine Microsoft Generative AI Hackathon—now done May 6, 2024

Do you privation to larn however to usage Azure AI and GitHub Copilot to physique multimodal apps that harvester text, image, video, oregon dependable inputs and outputs for a large impact? Do you privation to triumph currency prizes, Azure credits, and designation for your work? If you answered yes to immoderate of these questions, past you should articulation america for the Microsoft Generative AI Hackathon, an online situation wherever you tin usage your skills to make breathtaking generative AI solutions. 

Microsoft Build—May 21 to 23, 2024 

Develop the AI skills needed for tomorrow, today. At Microsoft Build 2024, you volition larn from in-demand experts, acquisition the latest innovations successful breakouts, and marque connections with a assemblage that tin assistance you execute more. Register for Build. 

Microsoft Developers AI Learning Hackathon, triumph up to $10,000 successful prizes! Now done June 17, 2024

This hackathon challenges you to propulsion the boundaries of what’s imaginable by gathering your precise ain AI app utilizing Azure Cosmos DB for MongoDB. Targeting some Node.js and Python developers, the hackathon guides you done broad learning to delve into the fundamentals of AI apps— and you mightiness triumph up to $10,000 successful prizes. By the extremity of the hackathon, you volition cognize however to physique your precise ain customized AI copilot with Azure Cosmos DB for MongoDB. Get started now! 

Azure Cosmos DB Conf 2024 disposable on-demand 

Missed the excitement and innovation-packed Azure Cosmos DB Conference successful April? No worries, we’ve got you covered! The lawsuit brought unneurotic the sharpest minds successful the industry, including merchandise leaders, main method officers, and assemblage members who shared insights connected gathering groundbreaking generative AI applications powered by Azure Cosmos DB. Whether you missed a small oregon missed it all, catch the replays on-demand. 

Microsoft’s AI Classroom Hackathon winners 

Microsoft’s AI Classroom Hackathon winners were chosen from much than 3,700 students from implicit 100 countries, participating successful Microsoft’s telephone to physique the adjacent procreation of intelligent applications that reimagine the aboriginal of acquisition utilizing Azure AI and Azure databases. Among the winners is Dialogues Through Time, that allows you to person interactive AI-driven dialogues with humanities figures similar Socrates and Leonardo DaVinci. Also, EquEdu, an AI-powered contented accessibility instrumentality that makes online acquisition contented accessible for the visually impaired. Lots of inspirational ideas here—check retired the winners and their projects. 

---

What’s new?  

Jessica shares insights connected exertion translation on with important updates and resources astir the data, AI, and integer exertion solutions that marque Microsoft Azure the level for the epoch of AI. Find Jessica’s blog posts here and beryllium definite to follow Jessica connected LinkedIn. 

The station What’s caller successful Azure Data, AI, and Digital Applications: Harness the powerfulness of intelligent apps  appeared archetypal connected Microsoft Azure Blog.