Enterprises automate deep research and analysis workflows at scale

Microsoft has announced the public preview of Deep Research in Azure AI Foundry, an offering that allows organizations to build programmable, composable AI agents for enterprise-scale research automation. The API and SDK-based platform integrates OpenAI's agentic research capabilities, brings web grounding with Bing Search, and is fully embedded in Azure's enterprise-grade agentic platform. Developers can automate complex research tasks, generate transparent, auditable outputs, and compose multi-step workflows across tools and agents. Deep Research is designed for integration with existing business applications and workflows, including Azure Logic Apps, and provides complete traceability and auditability for all research outputs—ideal for regulated industries. The system orchestrates a multi-step research pipeline leveraging the latest OpenAI models such as GPT-4 to clarify intent, scope tasks, ground research with Bing Search, and synthesize information into structured, compliant reports. Pricing is pay-as-you-go based on tokens. The preview is available for Azure AI Foundry Agent Service customers intending to embed research automation at the core of their digital transformation initiatives. The architecture enables triggering agents from anywhere—apps, workflows, or other agents—turning research into a reusable business service. Customer organizations can orchestrate multiple agents, automate reporting and notification, ensure compliance and observability, and ultimately embed research capabilities throughout their enterprise ecosystem. Initial industry interest is broad, spanning market analysis, regulatory reporting, analytics, and competitive intelligence, with a focus on security, flexibility, and integration potential.