CostLeak Exploratory analysis

CostLeak correlates a billing CSV with an optional activity CSV. It uses rule-based logic on cost and activity timestamps to classify resources as LEAKING, LIKELY LEAKING, or USEFUL. Results are indicative only. They depend on the accuracy and completeness of the data provided.

Files are processed ephemerally in memory and are not stored.

Resource ID Service Monthly Cost Annual Cost Status Confidence Recommendation

Frequently asked questions

What data is required?
A billing CSV with resource identifiers, service names, and costs. An activity CSV with resource identifiers and timestamps or usage metrics is optional but improves classification accuracy.
Is uploaded data stored?
No. Files are processed in memory only and discarded immediately after analysis. Nothing is persisted or logged.
Which cloud providers are supported?
Any provider. CostLeak accepts generic CSV inputs and infers common column names (e.g. resource_id, cost, last_activity). AWS, GCP, and Azure billing exports are typically compatible.
How accurate is the analysis?
Results are rule-based and indicative only. Accuracy depends on the completeness and correctness of your uploads. Confidence levels reflect the strength of available activity data.
Is this suitable for production decisions?
No. CostLeak is an exploratory tool. Use it to identify candidates for review, not as the sole basis for terminating resources or making financial commitments.

Indicative analysis only.

Privacy Policy · Terms & Conditions