AWS has launched AWS FinOps Agent in public preview. The agent-based AI solution automatically investigates cost anomalies down to their root cause and answers engineers’ cost questions directly in Jira and Slack.
A Cost Anomaly Detection alert signals that something has changed. The AWS FinOps Agent automatically takes it from there. It links cost changes to AWS CloudTrail events to determine who changed what and when, identifies the change that caused the spike, and generates an investigation report detailing the most likely cause and the responsible owner. Optionally, the agent opens a Jira ticket or posts a message in a Slack channel, so the engineer managing the resource can get the context immediately. Filters based on thresholds—for example, only anomalies above a certain amount—ensure that teams focus their attention on the most impactful changes.
Asking cost questions in natural language
Engineers can use AWS FinOps Agent to ask questions in plain language and receive answers based on their own cost and usage data. A question like “Why did my AWS costs increase last month?” yields an answer that identifies the cost changes, the services involved, and the underlying usage drivers. Using context files, such as account owner mappings, team definitions, and tag conventions, the agent tailors its responses to the organization’s terminology. The agent also remembers preferences across sessions.
The public preview also includes recurring cost reports in HTML, PDF, or PPT formats, as well as the ability to summarize recommendations from Cost Optimization Hub and Compute Optimizer in a Jira ticket.
AWS FinOps Agent is now available in the US East (N. Virginia) region and will be coming to our region at a later date. AWS has announced plans to expand the agent in the future to include cost analysis for AI workloads, among other features.
Tip: FinOps is shifting toward AI cost management and SaaS optimization