Amazon Web Services (AWS) today previewed an artificial intelligence (AI) agent specifically trained to both identify cost issues and surface recommendations for optimizing consumption of cloud infrastructure resources based on best financial operations (FinOps) practices.
Announced at the FinOps X 2026 conference, the AWS FinOps Agent has been trained to investigate the root causes of cost anomalies by reasoning across data pulled from AWS Cost Explorer, Cost Anomaly Detection, Cost Optimization Hub, and Compute Optimizer tools.
The AWS FinOps Agent then correlates the cost change with AWS CloudTrail events to identify the change that might have driven an increase in costs to generate an investigation summary that identifies the likely root cause and owner of the workload. Optionally, the agent can deliver the findings by opening a Jira ticket or posting to a Slack channel.
IT teams can also use natural language to surface cost and usage data specific to any workload they want to investigate. Additionally, IT teams can opt to schedule recurring cost reports to be rendered in an HTML, PDF, or PPT format.
Finally, IT teams can upload organization-specific context files that the AWS FinOps Agent will store in memory to provide any needed context across multiple sessions.
Jerry Rapisarda, director of engineering for AWS insights and optimizations, said the overall goal is to make it simpler for IT teams to use the insights generated by the AWS FinOps agent to either reduce costs or deploy additional workloads without increasing expense. IT teams can also integrate the AWS FinOps Agent with any automation framework they might be using to manage their workloads, he added.
The AWS FinOps Agent runs in the US East (N. Virginia) region of the AWS cloud but can be used to manage costs across AWS Regions and accounts when set up in the management account. During the preview period, the AWS FinOps Agent is available at no charge, with a monthly usage limit. Workday, AVIV Group, Convera, and Mitre 10 are among the early adopters of AWS FinOps Agent.
It’s not clear how much pressure IT teams are under to optimize consumption of cloud infrastructure resources. Historically, utilization rates of cloud infrastructure resources have been relatively low, which contributes to a significant waste of IT resources. However, during challenging economic times interest in IT infrastructure optimization tends to increase.
At the same time, many IT teams are now deploying AI workloads using expensive graphical processor unit (GPU) resources that, if not optimized, will slow the pace at which organizations are able to build and deploy these applications.
For now at least, the AWS FinOps agent doesn’t offer any insights into how much might be saved if an application were to be refactored, but hopefully as this type of data is shared more broadly application developers will make better architectural decisions.
In the meantime, however, most IT teams should assume that a significant portion of their existing budget dollars allocated to cloud computing resources could be better reallocated.
