A cloud efficiency workspace for safe decisions and governed automation
CloudKnife is built for multi-cloud environments: the same analysis and review pattern should span your clouds, even as we deepen support cloud by cloud. Today the strongest product depth is on Azure; AWS and GCP are on a quality-led roadmap, and we are open to early adopters who want to help shape what we build next. The workspace fits internal platform teams and MSPs who need repeatable, review-ready outputs across customer environments.
A review queue for evidence, context, and recommendations.
Working list of review-ready candidates with priority, types, and expected impact. Illustrative. Your tenant drives the real list.
The review workspace for serious candidates
Expand when you are ready. Impact, risk, rationale, resources, and Confidence stay together so reviewers decide with full context, whatever analysis produced the item.
CPU stays below the safety buffer, and tail behaviour remains stable in the decision window after review.
Ownership and policy boundaries stay visible for audit and review.
- Evidence and assumptions
- Impact and affected resources
- Safety headroom and risk context
Illustrative review surface. In product, comparison fields and policy context match your environment.
Your people still decide. CloudKnife removes the detective work that keeps efficiency stuck in the backlog.
What the platform evaluates
Two lenses, not four walls of cards. Everything below still attaches to serious candidates in product.
Production and spend reality
What actually ran and what it cost, not a single averaged line.
- Utilisation, seasonality, and bursty work that averages flatten away.
- Expected impact with assumptions explicit enough to challenge.
- Trade-offs surfaced so reviewers compare options in one place.
Ownership, policy, and blast radius
Who answers for change and which rules apply before automation is even offered.
- Workload shape, environment tags, and ownership travel with each candidate.
- Production versus non-production expectations kept visible.
- Guardrails for critical tiers, windows, and blast radius.
- Review expectations and automation limits before anything runs.
From signal to governed action
A straight path from observation to review, with automation only where policy and evidence support it.
Ingest usage, configuration, and environment context with read-only access first where that applies.
Cross-check behaviour, cost, risk, and ownership so candidates are explainable, not guessed.
Package rationale, impact, resources, and Confidence into a review-ready item.
Teams approve what goes live. Policy-governed automation follows only where you allow it, not by default.
What teams can act on
The same review and governance pattern applies across these families at the platform level. A dedicated walkthrough covers optimisation prioritisation, queue behaviour, and reviewer mechanics step by step.
Rightsizing
Match capacity to observed demand while keeping safety headroom explicit. Recommendations carry rationale, expected impact, affected resources, and Confidence so production stakeholders can align before anything changes or automates.
- Scheduling
Align runtime to predictable patterns with clear impact expectations before anyone opts in.
- Hygiene and cleanup
Surface idle or orphaned resources with accountability so owners confirm intent before change.
- Service-fit improvements
Highlight better-fitting services or SKUs when constraints and workload needs point that way.
Multi-cloud direction, Azure-first depth today
We add clouds only when recommendation quality and review clarity match what operators need in production. Microsoft Azure is the supported starting point. AWS and Google Cloud are planned next on the same bar. Dedicated sovereign or national variants are not supported today. Teams on AWS or GCP are welcome to talk to us about early design partnership.
Microsoft Azure is supported today: read-only onboarding first, insights, and review-ready recommendations with context.
AWS and Google Cloud are planned next under the same quality bar, not as noisy signal lists.
Sovereign or national variants stay out of scope until we can validate them properly.
Why teams trust it
Operator control first. Evidence in the open. Automation earned through policy, not assumed.
Rationale, impact, resources, and Confidence stay attached so decisions are inspectable.
Automation is bounded by rules and review history, not silent execution across your cloud.
Approvals and review context stay tied to each item so audits stay grounded in evidence.
CloudKnife earns trust in review first. It does not promise unattended execution everywhere today.
How it improves over time
CloudKnife learns from approvals, rejections, and operating patterns so recommendations align with how your team actually decides.
The system carries forward what “yes”, “no”, and “not yet” mean in your environment. That feedback tightens prioritisation and policy matching, without extra dashboard housekeeping.
Next step
See the platform where we support you today
Request access for a short conversation, open the optimisation walkthrough for queue and reviewer detail, or talk MSP delivery. Deepest onboarding today is on Azure; ask if you are on AWS or GCP and want to explore what comes next.