Google built an engine for waiting AI agents. Customer preview comes later.
Google Cloud described Agent Compute Engine on 22 October as infrastructure designed for agents that spend time waiting for tools or human approval.
The company says it already uses the technology internally and expects a customer preview in the coming months.
Google describes isolated virtual machines that can pause and resume. The proposed benefit is to avoid keeping the same compute resources active throughout a workflow’s idle periods.
The announcement also describes handling credentials outside the agent’s virtual machine. It does not give a customer launch date or a measured saving that every workload can expect.
Our reading: a useful trial would measure active work separately from waiting time, using a workflow that really needs approval. A pause-and-resume design may suit that pattern, but its value for a particular business process still needs to be demonstrated when the customer preview becomes available.
Sources
Our file on Google
- 29 Oct
Google Cloud's sales doubled to $30.7 billion in a quarter. Alphabet's profit still fell to $26.2 billion from $35 billion.
- 27 Oct
Microsoft rushed out a .NET security fix on Sunday. Google's Brotli code was patched in February, but no one saw the danger.
- 22 Oct
Google Cloud lost one London zone. HSBC users reported problems.
- 20 Oct
Cognizant assigned 18,000-plus staff to build Google AI agents for clients. Terms were undisclosed.
- 15 Oct
AWS and Google's lobby priced EU cloud rules at €115bn. Brussels estimated €18bn.
For your next steering committee
- 1Ask which AI tools already touch your ERP or finance data, including the ones nobody approved.
- 2Ask what each tool is allowed to change, and who reviews what it changed.
- 3Ask how the result is measured, and who decides to switch a tool off.
Issue 01 of the ERP LEADERS brief puts a timeline, the three-number review and six questions on one page. Read issue 01
Welcome back. · Issue 01
