Posts

Amazon Bedrock AgentCore Browser now supports proxy configuration - devamazonaws.blogspot.com

Amazon Bedrock AgentCore Browser now supports customer-provided proxy configuration, enabling customers to route browser sessions through their own proxy infrastructure for geo-targeting, regional content access, and compliance requirements. Organizations can access geo-restricted content, verify region-specific pricing, and validate localized application behavior across markets. Customers in regulated industries such as healthcare and financial services can route traffic through corporate proxy infrastructure to meet security policies while automating critical business processes. Browser proxies help eliminate re-authentication cycles due to rotating IPs, while providing stable, controllable egress addresses for IP allowlisting requirements. The feature currently supports both HTTP and HTTPS protocols with secure credential management through AWS Secrets Manager. This enhancement is particularly valuable for healthcare organizations accessing portals with strict IP allowlisting, fin...

[MS] Handling Legacy User Settings in SharePoint Framework - devamazonaws.blogspot.com

Introduction The SharePoint Framework (SPFx) has undergone some important changes in how it handles Entra ID app registration and security models. Here you can find additional details about what was updated and about the new application IDs used by SPFx. These changes have important implications for developers who previously relied on legacy user settings stored through the old application registration model. This article explores the technical challenges and provides practical solutions for handling legacy user settings when transitioning to the new SharePoint Framework security architecture. Understanding the SharePoint Framework Security Model Evolution The Legacy Application Model Previously, SharePoint Framework utilized an application called SharePoint Online Client Extensibility Web Application Principal . This application had several key characteristics: Tenant-specific registration : The application was automatically created in each customer tenant when they began using S...

[MS] Build Intelligent Apps with SQL: Join the SQL + AI Datathon - devamazonaws.blogspot.com

Image
The SQL + AI Datathon is a hands‑on challenge designed to show how the foundations for building modern, intelligent applications with SQL. Over a set of guided missions and a focused open hack, you’ll learn how to combine SQL with embeddings, semantic search, and Retrieval Augmented Generation (RAG) to build real AI‑powered experiences. The SQL + AI Datathon puts SQL at the center of the architecture. You’ll learn how to: Use SQL as a reliable, secure data source for AI workloads Implement RAG (Retrieval‑Augmented Generation) so AI responses are grounded in your data Build AI experiences that scale from prototypes to real applications Who Should Participate? The SQL + AI Datathon is designed for: Developers curious about AI Developers building AI applications with real data Anyone interested in trying out the new AI Features in SQL Server 2025 Any beginner thinking about how to get started with data powered AI applications for free Learn Along the Way with the Reactor Se...

[MS] How we built the Microsoft Learn MCP Server - devamazonaws.blogspot.com

Image
When we launched the Microsoft Learn Model Context Protocol (MCP) Server in June 2025, our goal was simple: make it effortless for AI agents to use trusted, up-to-date Microsoft Learn documentation. GitHub Copilot and other agents are increasingly common, and they need to be able to ground responses just like humans with browsers do. Learn MCP Server is a remote server that exposes agent-friendly tools over Streamable HTTP Transport, backed by the same Learn knowledge service described in How we built “Ask Learn” . Why MCP and why Learn MCP Server? Modern AI agents can discover and use tools dynamically through Model Context Protocol , a standard that lets agents negotiate capabilities at runtime, stream results, and adapt as tools evolve. Instead of manually searching, scraping pages, or maintaining embeddings, clients simply connect to Learn MCP Server, request the list of tools available and invoke tools as the agent decides. VS Code GitHub Copilot and other MCP Clients can deli...

[MS] Beyond the Prompt - Why and How to Fine-tune Your Own Models - devamazonaws.blogspot.com

Image
Large Language Models (LLMs) have reached a point where general intelligence is no longer the bottleneck. The real challenge in enterprise AI systems behavioral alignment ensuring models that produce consistent, reliable, policy-compliant outputs on a scale. Prompt engineering and Retrieval-Augmented Generation (RAG) are powerful but they do not change model behavior. Fine-tuning will solve this by customizing a pretrained AI model with additional training on a specific task or dataset to improve performance, add new skills, or enhance accuracy. This post explores what Microsoft Foundry fine-tuning is, when using it, the fine-tuning approaches it supports   and code examples on how to run Fine-tuning on Microsoft Foundry. What Is Microsoft Foundry Fine-Tuning: Microsoft Foundry fine-tuning allows you to customize pre-trained foundation models (OpenAI and open models) using task-specific datasets, producing a specialized model that behaves predictably for your use case while mai...

[MS] Deploy to Azure App Service deployment slots with azd - devamazonaws.blogspot.com

Deploy to Azure App Service slots with azd deploy Missed deploying to an App Service slot? Now you can—without workarounds or manual scripts. What's new? Before, azd deploy could only deploy to the default production slot. Now, azd understands the slots you define in Bicep and deploys to the slot you designate. On first deploy, azd deploys to both production and any defined slots. After that, if you have more than one slot, it prompts you to choose—or you can set the slot with an environment variable. Why it matters Slot deployment is essential for blue/green, staging, and testing flows, but it wasn't possible from azd before. Now, you can go from Bicep definition to deploying code to a slot—no extra scripts, extensions, or guessing where your code lands. It's a smoother, more predictable experience using just azd . How to use it Define a slot in your Bicep: resource webApp 'Microsoft.Web/sites@2021-03-01' = { name: 'my-appservice' location: r...

[MS] Maximize Azure Cosmos DB Performance with Azure Advisor Recommendations - devamazonaws.blogspot.com

Image
In the first post of this series, we introduced how Azure Advisor helps Azure Cosmos DB users uncover opportunities to optimize efficiency and make smarter decisions. This follow-up dives deeper into one of the most important categories of guidance: performance. If you've ever dealt with skewed workloads, unexpected RU consumption, or queries that don’t scale as expected, Azure Advisor surfaces exactly the insights you need to diagnose and improve them. In this post, you’ll learn what performance recommendations Advisor provides, why they matter, and how you can use them to enhance the speed, reliability, and scalability of your workloads. Exploring Azure Advisor Performance Recommendations Below are some of the key performance related recommendations Advisor provides for Azure Cosmos DB, along with guidance you can use immediately. Use hierarchical partition keys for optimal distribution: Poor partitioning is one of the most common causes of hot partitions , throttling (429 ...