Arize AI showcased closer collaboration with Microsoft and a series of new integrations at the Microsoft Ignite conference. These updates come at a crucial time for enterprise adoption of generative AI. Although most companies are planning production deployments of LLMs, more than half still cite issues such as data privacy and difficulty identifying problems (i.e. accuracy of answers) as obstacles that prevent them from moving from prototype to successful real-world deployment.
Azure AI’s advanced AI and cloud capabilities, along with Arize’s deeply integrated platform, help enterprise teams – including airlines, financial services companies, retailers and enterprises Fortune 500 technology companies – to overcome these obstacles. With Arize integrated directly into Azure AI Studio and available in Azure Marketplace, enterprise AI teams can easily launch Arize’s essential LLM assessment and observability tools while keeping data within Azure. In addition to this easy integration, the Arize platform has several new integration points with Azure.
Here are some of them: Azure AI App Template Gallery: Azure developers now have access to pre-built Arize templates to help monitor LLM models and applications, troubleshoot tool extraction and execution issues, and improve the RAG pipeline – all in a gallery regularly tested and used by thousands of developers around the world, with built-in security including keyless authentication to reduce the risk of API key leaks. Azure Native Integrations: Announced in private preview, developers now have an integrated development experience in the Azure portal, SDK and CLI for Arize AI with integrated identity management and transparent billing. Azure AI Model Catalog: The Azure AI Model Catalog offers several paid inference APIs through Models-as-a-Service (MaaS); Arize’s integration with Azure MaaS allows teams to objectively assess LLM systems and trace complex issues.
