Investigate a cost anomaly end-to-end. Triggers on phrases like "cost spike", "unexpected spend", "anomaly", "what's driving this cost". Pulls billing data, identifies the top contributing SKUs/projects/labels, drafts a hypothesis about cause, and produces both the engineering Slack message and the CFO digest paragraph in house style.
Report on current utilization across all active commitments (CUDs, RIs, Savings Plans). Triggers on phrases like "commitment utilization", "are we using our CUDs", "RI report", "savings plan check", "commitment portfolio". Computes current utilization, days remaining, projected over- or under-use, and the dollar impact of any gap.
Generate a standardized writeup after a cost anomaly has been resolved. Triggers on phrases like "postmortem for the X spike", "writeup for the anomaly", "document what happened with X". Produces a structured document covering timeline, root cause, dollar impact, what was changed, and prevention — so postmortems are comparable across the team over time.
Produce a one-paragraph executive summary on a FinOps topic, in house style. Triggers on phrases like "exec digest", "CFO summary", "one-paragraph summary for leadership", "write this up for the exec readout". Lead with dollars, no jargon, action-oriented close.
Prepare the FinOps section of the monthly business review. Triggers on phrases like "MBR prep", "monthly business review", "monthly cost review", "I need to put together my MBR". Pulls last month's spend by service and team, computes trends vs the prior 3 months, drafts the narrative that goes in the deck, and surfaces the anomalies leadership will ask about.
Save and manage prompts that worked well. Triggers on phrases like "save this prompt", "add to prompt library", "this is a keeper", "remember this prompt". Captures the prompt with metadata (use case, model version, what it's good at, what to watch for) so it can be retrieved and reused later.
Generate ticket-ready right-sizing actions for a service, workload, or team. Triggers on phrases like "right-size", "rightsizing", "are we over-provisioned", "what can we shrink", "utilization opportunities". Pulls utilization data, generates concrete recommendations with dollar impact estimates, formatted so the practitioner can paste each one into a ticket and assign it.
Compute tag coverage for a project, org, or business unit, and draft the cleanup communications. Triggers on phrases like "tag coverage", "tagging compliance", "find untagged resources", "who isn't tagging properly". Outputs per-tag-key compliance percentages, lists of non-compliant teams/services, and ready-to-send Slack DMs to the likely owners.