Advanced AI Technologies

Google advances persistent enterprise AI agents

Google introduced a persistent Gemini enterprise agent that can retain context, delegate subtasks, and route work across models.

Google introduced a persistent Gemini agent for enterprise work, according to a weekly AI developments report. The agent is described as operating in the cloud for extended periods while retaining context, delegating subtasks, and routing work between Google’s Gemini models and Anthropic’s Claude. Additional models were expected to be supported. The system represents a shift from short, user-triggered prompts toward longer-running enterprise agents that can manage multi-step assignments. Persistent context and delegated subtasks could support research, document workflows, coding, analysis, and other knowledge-work processes. Model routing may allow tasks to be matched with different systems according to capability or operational requirements. The approach also raises practical needs around permissions, monitoring, security, and human approval because an agent that continues working over time can affect more systems and data than a conventional chatbot. The announcement reflects intensifying competition among cloud and enterprise vendors to provide agent platforms rather than standalone generative-AI interfaces.

Original source: Open Data Science

Related RAQI solutions

In-depth analysis on RAQI Insight

Work with RAQI