Google previews Gemini Enterprise agent for work across apps and devices

Google announced a workplace Gemini agent at its Gemini at Work event on October 8, available to enterprise customers in private preview. Within the Gemini Enterprise app, users can chat with the agent and assign tasks through one interface. Google says it can work inside Gmail, Drive, Docs, Sheets, and Calendar, alongside third-party apps such as Slack and Microsoft 365. The cloud-based agent maintains context across connected devices, coordinates job-specific sub-agents, and selects an AI model suited to each task. It can also operate as a coworker agent with a dedicated identity and an email address under @agents.company.com.

The market significance is Google's attempt to make the enterprise agent interface a coordination layer across workplace software. Persistent context and task delegation address the orchestration question in agentic systems: whether value accrues to individual assistants or to the system that connects them and carries work between applications. Workspace integration gives Google an existing workplace surface, while the reported third-party connections extend the product's scope. Private-preview status leaves production reliability and customer adoption unproven; the announcement establishes product direction, not demonstrated enterprise returns.

For builders and investors, a concrete diligence question is how reliably delegated tasks complete across application boundaries. Evaluation should examine context continuity, sub-agent handoffs, and the permissions attached to a dedicated agent identity. Specialist agent vendors will need to show measurable workflow value within an environment where Google proposes to coordinate multiple tools from one interface. The source provides no pricing or customer outcome figures to assess that value yet.