Microsoft Foundry - 5

2026 (6)

June (4)

ai: Foundry IQ: Build smarter agents faster with unified knowledge and serverless retrieval

Microsoft AI Foundry Blog

2026-06-02

Foundry IQ is presented as a way to ground agents in unified enterprise knowledge using serverless retrieval. The post emphasizes improved agentic retrieval quality and production-ready security.

ai: Build agents you can trust across any framework with open evals and a control standard

Microsoft AI Foundry Blog

2026-06-02

Microsoft describes a set of tools for building AI agents with open evaluations, portable runtime controls, observability, and security workflows that are intended to work across frameworks. The post focuses on trust and control for agent development rather than a single framework-specific feature set.

ai: What’s new in Microsoft Foundry | Build Edition

Microsoft AI Foundry Blog

2026-06-02

Microsoft Build 2026 introduces several Microsoft Foundry updates for agent developers, including hosted runtimes, Toolboxes, memory, Voice Live, Foundry IQ, new models, managed compute, and added trust, evaluation, and observability features.

ai: What’s New in Hosted Agents in Foundry Agent Service

Microsoft AI Foundry Blog

2026-06-02

Microsoft announced new updates for hosted agents in Foundry Agent Service at Build, focused on easier deployment, broader modality support, and tighter integration for enterprise use. The article is an overview of recent hosted agent runtime improvements.

May (2)

ai: What’s new in Microsoft Foundry | May 2026

Microsoft AI Foundry Blog

2026-05-31

May 2026 Microsoft Foundry updates include trace-based evaluation for any agent on any cloud, new catalog models (Grok 4.3 and DeepSeek V4), GPT-5 Reinforcement Fine-Tuning at gated GA, and several platform and SDK additions. The post also notes GA for Managed VNET and Content Understanding improvements, plus Foundry Local updates and a new azure-ai-projects release.

ai: How to run evals for the model router

Microsoft AI Foundry Blog

2026-05-19

The post explains how to evaluate the Foundry model router using an open-source GitHub repo for router-aware eval pipelines. It covers running quality, cost, and latency evaluations.