Question 1

What is CloudKeeper?

Accepted Answer

CloudKeeper is the CloudThinker home for CostOps — the agentic execution layer above FinOps. It runs a daily eight-phase loop across AWS, GCP, Azure, and Kubernetes (detect, isolate, trace, wash, chase, open MR, ship under approval, learn), and lands every change as a Merge Request with the diff, rationale, projected savings, and rollback.

Question 2

How is CloudKeeper different from rule-based FinOps tools?

Accepted Answer

Rule-based tools surface a flood of recommendations against static thresholds; humans triage. CloudKeeper reads workload context, verifies against live state, and only ships when the change is safe and material. The output is a landed Merge Request, not a dashboard row. The false-positive flood is gone.

Question 3

Does CloudKeeper cover AI and LLM spend?

Accepted Answer

Yes. Token usage is a first-class signal alongside compute and storage. CloudKeeper attributes AI spend back to features, services, and teams; surfaces anomalous prompts (looped agents, runaway batch jobs); and proposes model-selection changes when a cheaper tier would meet the SLA.

Question 4

How does CloudKeeper compare to Cost Explorer, CloudHealth, Cloudability, Vantage, or Kubecost?

Accepted Answer

CloudKeeper differs at the unit of work. Cost Explorer, CloudHealth, Cloudability, Vantage, and Kubecost stop at the dashboard recommendation. CloudKeeper ships the change as a Merge Request under the approval gate the team chose, and attributes realized savings back to the original anomaly so the loop closes on data, not on quarterly reviews.

Turn cloud waste into safe, continuous savings.