Azure AI Data, AI, and Digital Apps updates:
Modernize data, build smart apps, and use AI
Generational AI models and tools have improved application and business process experiences for years, but this year was a turning point. Within months, customers and partners integrated AI into their transformation roadmaps and launched AI-powered Digital Apps and services.
A new technology has never caused such rapid change. It shows how many organizations were AI-ready and how cloud, data, DevOps, and transformation cultures prepared them. Customers and partners can maximize AI with hundreds of Microsoft resources, models, services, and tools this year.
New models and multimodal capabilities in Azure AI
They offer the most advanced open and frontier models so developers can build confidently and unlock immediate value across their organization.
They added Models as a Service to Azure OpenAI Service last month. Azure AI applications can use model providers’ latest open and frontier LLMs.
MaaS for Llama 2 was announced last week. The ready-to-use API and token-based billing of MaaS for Llama 2 let developers integrate with their favorite LLM tools like Prompt Flow, Semantic Kernel, and LangChain. Hosted fine-tuning lets generative AI developers use Llama 2 without GPUs, simplifying model setup and deployment. Llama 2, purchased and hosted on Azure Marketplace, lets them sell custom apps. In his blog post, John Montgomery describes this announcement and shows Azure AI Model Catalog models.
Here are improving Azure OpenAI Service and launched multimodal AI to let businesses build
- Generative AI experiences with image, text, and video:
- Preview DALL·E 3: Generate images from text descriptions. The AI model DALL·E 3 excels in this regard. DALL·E 3 generates images from user descriptions.
- General availability: GPT-3.5 Turbo preview at
- 16k token prompt:
- GPT-4 Turbo: Azure OpenAI Service models now extend prompt length and improve generative AI application control and efficiency.
- Visionary GPT-4 Turbo preview: GPT-4V optimises experiences by generating text output from images or videos using Azure AI Vision enhancements like video analysis.
- Modifying Azure OpenAI Service models: Fine-tune Azure OpenAI Service models Babbage-002, Davinci-002, and GPT-35-Turbo. Developers and data scientists can customize Azure OpenAI Service models. Discover fine-tuning.
- GPT-4 updates: Azure OpenAI Service can fine-tune GPT-4. Organizations can customize the AI model by fine-tuning. It’s like customizing an AI suit. Checking GPT-4 updates.
- Frontier steering: Prompting power grows
GPT-4 prompting dazzles! Microsoft Research recently blogged about promptbase, a reasoning-based GPT-4 prompt. Other AI models lag behind GPT-4 in various test sets, including those used to benchmark the recently announced Gemini Ultra. Zero-shot chain-of-thought prompting. See the blog post and try these GitHub resources.
LLMOps RAI tools and practices
Safety boundaries supporting short- and long-term ROI must be considered as AI adoption matures and companies produce AI apps. This month’s LLMOps for business leaders article covered integrating responsible AI into your AI development lifecycle. These Azure AI Studio best practices and tools help development teams apply their principles.
AI Advantage from Azure
Cloud database Azure Cosmos DB uses AI. Built-in AI, natural language queries, vector search, and simple Azure AI Search integration are supported. To demonstrate these benefits, her new Azure AI Advantage offer gives new and existing Azure AI and GitHub Copilot customers 40,000 RU/s of Azure Cosmos DB for 90 days.
Modern data and analytics platforms are essential for AI transformation because intelligent apps need data.
Multinational law firm Clifford Chance benefits clients with new technology. The company built a solid Azure data platform to test Azure AI, Microsoft 365 Copilot, and large language models. Cognitive translation is an IT team’s fastest-growing product.
Azure Machine Learning and Azure Databricks helped Belgian insurance company Belfius reduce development time, efficiency, and reliability. Data scientists can create and transform features while the company improves fraud and money laundering detection.
Co-innovation Azure-Databricks improves AI experiences
Customers and partners shared how maturing AI tools and services are helping them achieve more at Microsoft Ignite 2023 in November.
One of her strategic partners, Databricks offers Azure’s fastest-growing data services. Azure Databricks’ interoperability with Microsoft Fabric and use of Azure OpenAI to improve customer AI experiences were recently demonstrated. Customers can build retrieval-augmented generation (RAG) applications on Azure Databricks and analyze the output with Power BI in Fabric using Azure OpenAI LLMs.
Azure Database PostgreSQL AI extension
With the new Azure AI extension, Azure OpenAI LLMs can generate vector embeddings and build rich PostgreSQL generative AI applications. These powerful new capabilities and pgvector support make Azure Database for PostgreSQL another great place to build AI-powered apps.
SQL Server anywhere gets Azure Arc cloud innovation
SQL Server manageability and security improvements from Azure Arc are available this month. Customers can optimize database performance and gain critical SQL Server estate insights with SQL Server monitoring. Azure portal makes Always On availability groups, failover cluster instances, and backups more visible and simple.
Lower Azure SQL Database prices Calculate hyperscale
The new Azure SQL Database price Hyperscale gives cloud-native workloads Azure SQL performance and security at commercial open-source database prices. For scalable, AI-ready cloud applications of any size and I/O, hyperscale customers can save 35% on compute resources.
Apps change operations and experiences
Personalized employee apps and customer chatbots are examples of digital applications developed and deployed by companies using AI to improve operations and experiences. Updates like these enable innovation.
Custom copilots and the seven AI development pillars
Copilots are exciting, and Azure AI Studio in public preview lets developers build generative AI apps. Businesses must carefully design a durable, adaptable, and effective approach for this new era. What can AI developers do for customer engagement? Consider these seven pillars for your custom copilot.
AKS is a top cloud-native intelligent app platform
AI and Kubernetes will influence app development. AI and cloud-native collaboration drives innovation at scale, with Azure Kubernetes Service (AKS) supporting compute-intensive workloads like AI and machine learning. Brendan Burn’s KubeCon blog describes how Microsoft builds and supports customer-beneficial open-source communities.
- Azure enables unlimited innovation.
- Her recent portfolio news, resources, and features, especially digital applications, have received great tech community and customer response.
- Ignite’s Platform Engineering Guide is a hit, demonstrating demand for this training.
- Technology innovation in companies is crucial.
- Two recent customer stories caught my eye.
- Modernizing LEGO House interactive experiences with Azure Kubernetes
Here helping The LEGO House in Denmark, the ultimate LEGO experience center for kids and adults, migrate custom-built interactive digital experiences from an aging on-prem data center to Microsoft Azure Kubernetes Service (AKS) to improve stability, security, and iteration and collaboration on new guest experiences. LEGO House updates experiences faster with this cloud move and guest feedback. The modernizing destination hopes to share knowledge and technology with LEGOLAND and brand retail stores.
Gluwa chose Azure for a reliable, scalable cloud solution to bring banking to emerging, underserved markets and close the financial gap.
An estimated 1.4 billion people struggle to get credit or personal and business loans in a country with limited financial infrastructure. Borderless financial technology from blockchain helps Gluwa with Creditcoin stand out. The Azure cloud supports it. Gluwa has a solid platform with her.NET framework, Azure Container Instances, AKS, Azure SQL, Azure Cosmos DB, and more. The business is more efficient due to reliable uptime, stable services, and rich product offerings.
CARIAD builds Volkswagen Group vehicle service platform with Azure and AKS
The Volkswagen Group’s software subsidiary CARIAD created the CARIAD Service Platform with Microsoft using Azure and AKS to provide automotive applications to Audi, Porsche, Volkswagen, Seat, and Skoda as the industry moved to software-defined vehicles. This platform lets CARIAD developers develop and service vehicle software, giving Volkswagen an edge in next-generation automotive mobility.
AKS and Azure Arc help DICK’S Sporting Goods provide omnichannel service
To create a more consistent, personalized customer experience across its 850 stores and online retail experience, DICK’S Sporting Goods envisioned a “one store” technology strategy to write, deploy, manage, and monitor its store software across all locations nationwide and reflect those experiences on its eCommerce DICK’S needed modularity, integration, and simplicity to integrate its public cloud and edge computing systems.
DICK’s Sporting Goods is using Azure Arc and Azure Kubernetes Service to migrate its on-premise infrastructure to Azure and create an adaptive cloud environment. The retailer can now easily deploy new apps to every store for ubiquity.
Performance and efficiency of Azure Cobalt for intelligent apps
Azure has hundreds of cloud-native and intelligent application performance services. Azure silicon performance and efficiency efforts have expanded. Azure Maia, her first custom AI accelerator series for cloud-based AI training and inference, and Azure Cobalt, her first Microsoft Cloud CPU, were launched recently.
Azure Arm chips perform 40% slower than Cobalt 100, the first 64-bit 128-core chip in the series, which runs Microsoft Teams and Azure SQL.
0 Comments