Azure AI Foundry

Short Intro: Azure AI Foundry is Microsoft's consolidated AI development platform, announced at Microsoft Ignite in November 2024 as the merger of five previously separate services: Azure OpenAI Service, Azure AI Studio, Azure Machine Learning, Azure Cognitive Services, and Azure AI Search. Built by Microsoft and backed by an $80B datacenter investment commitment in 2025, the platform gives enterprise engineering teams access to 1,900+ foundation models alongside a managed Agent Service, long-term memory infrastructure, and deep integration with GitHub, Visual Studio Code, Microsoft 365, and Azure Logic Apps. More than 10,000 organizations including Heineken, Fujitsu, and Carvana have migrated workloads onto the platform.

Key Capabilities:

• Model catalog with 1,900+ foundation models including GPT-4o, GPT-5, Phi, Llama, Mistral, and DeepSeek R1

• Azure AI Foundry Agent Service for building, deploying, and scaling enterprise AI agents

• Agent Memory Service with long-term context extraction, consolidation, and retrieval across sessions

• Open-source SDK in Python, C#, and JavaScript with model switching without recoding

• Prompt Flow pipeline orchestration for multi-step AI workflows

• Azure AI Search for RAG across SharePoint, Azure Fabric, and Bing

• Fine-tuning on serverless and managed compute with Responsible AI guardrails applied uniformly

• Azure AI Content Safety with multi-modal detection across text, images, and voice

• Microsoft Purview integration for automatic classification of data accessed by agents

• 1,400+ Azure Logic Apps connectors for workflow triggers and integrations

• Microsoft 365 Agents SDK for deploying agents across 15+ Microsoft channels

• GitHub and VS Code integration for model and agent management without portal context-switching

• Model evaluation framework for quality testing before production deployment

• Budget controls and token consumption monitoring for cost governance

See Azure AI Foundry pricing details →