Enterprise agents scale

Google Cloud and Databricks are pushing AI agents deeper into workplace software, turning chat assistants into tools that can take actions inside apps. (cloud.google.com) (databricks.com) At Google Cloud Next on April 22, Google introduced the Gemini Enterprise Agent Platform and said Gemini Enterprise now includes more than 90 partner agents from companies including Adobe, Atlassian, Salesforce, ServiceNow, Workday, Oracle and Palo Alto Networks. (cloud.google.com) (googlecloudpresscorner.com) Google’s documentation says Gemini Enterprise is designed to manage agents from Google, third parties and internal teams in one place, with centralized oversight for what those agents can access and how they are deployed. (docs.cloud.google.com 1) (docs.cloud.google.com 2) Databricks, in posts published April 15 and April 17, said its Unity AI Gateway can connect agents to external Model Context Protocol servers for tools including GitHub, Atlassian and Glean, while routing permissions and audit controls through Unity Catalog. (databricks.com 1) (databricks.com 2) Model Context Protocol, or MCP, is a standard way for an AI agent to call outside tools, like a universal plug shape for software. Google and Databricks are both packaging that access with admin controls, logs and identity rules so companies can let agents act without handing them broad, untracked permissions. (databricks.com) (docs.cloud.google.com) The shift is moving from assistants that answer questions to agents that can open tickets, read documents, pull code context and trigger workflows across systems employees already use. Google said Gemini Enterprise already connects to applications such as Jira, Confluence, Microsoft SharePoint and ServiceNow through permissions-aware search and agent features. (docs.cloud.google.com) (cloud.google.com) Google and Salesforce used the launch to show that cross-vendor integrations are becoming part of the pitch. Salesforce said on April 22 that customers will be able to deploy agents in Slack and Google Workspace, with Agentforce and Gemini Enterprise sharing context across the two companies’ systems. (salesforce.com) (cloud.google.com) Databricks is framing the same problem from the governance side, especially for software teams using multiple coding agents. Its Unity AI Gateway documentation says the service covers large language model endpoints, MCP servers and coding agents, with usage analysis, permissions, guardrails and capacity controls across providers. (docs.databricks.com) (learn.microsoft.com) The contest is no longer just over whose model writes the best answer. It is increasingly over which platform becomes the control layer for agents working inside the ticketing, design, coding and customer-service systems companies already pay for. (cloud.google.com) (databricks.com)