CloudKnife

Cloud teams tend to run oversized infrastructure because under delivery pressure, adding headroom is the safe move and nobody has time to prove a change won't hurt. Existing tools like Azure Advisor only surface potential savings from usage data and hand engineers a list of individual recommendations to investigate, without workload context, without combining related actions, and without accounting for existing discounts or commitments. This leaves teams to manually organise and implement changes, so savings opportunities (oversized VMs, idle resources, orphaned disks, unused commitments, old snapshots) pile up and go unaddressed.

CloudKnife consolidates cloud cost and resource data in one place, explains cost evolution and drift with root causes, and slices spend by application, team or customer. It analyzes compute, databases, storage and commitments in context, factoring in workload role, usage patterns and existing discounts, to produce consolidated recommendations (resizing, scheduling, reservations, cleanup) with evidence and reasoning rather than a raw list. Changes start as proposals that users review and approve, and an opt-in 'Autopilot' feature can continuously adjust infrastructure within user-defined policies and IaC guardrails. It currently supports Microsoft Azure, with Google Cloud and AWS listed as upcoming.