News
Microsoft's Azure AI Foundry: New Tools for Streamlined AI Development and Operations
Microsoft has introduced Azure AI Foundry, a comprehensive platform designed to consolidate its generative AI tools under a unified system. Announced during the opening day of the Ignite 2024 conference, the platform aims to simplify the deployment and management of generative AI applications through an intuitive interface and centralized tools.
"Now, AI development, operations, and compliance teams can easily create, manage, and audit their organization's hubs, projects, and resources from within the Azure AI Foundry portal, reducing the need to visit Azure Portal or different areas of Azure AI Foundry portal for day-to-day administrative tasks," Microsoft stated in blog post.
The core components of Azure AI Foundry include the Foundry Portal (formerly Azure AI Studio), the Foundry SDK, and a range of tools like AI Agents and enhanced Retrieval-Augmented Generation (RAG) search capabilities. These features are tailored to optimize every stage of the AI development lifecycle, from model training to deployment and monitoring, ensuring organizations can innovate while maintaining compliance and efficiency.
Microsoft is touting many new and enhanced features arriving with the new platform, including:
- Continuous online evaluation: This feature allows organizations to assess application outputs using built-in or custom metrics. It supports various deployment methods, enabling early detection of quality and safety issues during both pre-production and production stages
- Custom dashboards in Azure Monitor: Azure Monitor allows teams to visualize performance trends, such as token usage, user feedback and evaluation results. Customizable dashboards provide tailored metrics and streamline incident response.
- Enhanced observability with Tracing: Tracing capabilities, available via the Azure AI Foundry SDK, provide detailed insights into application workflows. Developers can monitor inputs, intermediate results and outputs while tracking execution metrics for optimization and debugging. Logs are accessible in Azure AI Foundry or Azure Monitor Application Insights.
The company said that its Azure AI search, now found in Azure AI Foundry, including enhanced Retrieval-Augmented Generation (RAG) capabilities, which Microsoft said will improve results by 12.5 percent. It will also be increasing its GitHub integration search to provide users with better results.
A bigger inclusion into Azure AI Foundry is the newly announced Azure AI Agent Service, which will allow enterprises to automate workflows through orchestration lines. IT sets the task, and the new service will handle the task independently before handing the task off for final human review. Microsoft said the new service will leverage across many data points across a network.
"A standout feature of Agent Service is the ability to easily connect enterprise data for grounding, including Microsoft SharePoint and Microsoft Fabric, and tools integration to automate actions," wrote Microsoft. "With features like bring your own storage (BYOS) and private networking, it ensures data privacy and compliance, helping organizations protect their sensitive data. This allows your business to leverage existing data and systems to create powerful and secure agentic workflows."