Kubernetes costs — finally clear

OneLens gives platform and engineering teams a complete, correlated view of Kubernetes costs — from cluster to workload — across EKS, AKS, and GKE.
See what you're spending, where it's wasted, and who should own it.

Trusted by leading businesses

The problem with Kubernetes cost visibility

Cloud bills don't map to Kubernetes. Native tools show compute costs at the instance level, but they can't tell you what a specific workload actually costs, how efficiently it's using resources, or which team should be accountable for the spend.

OneLens bridges that gap.

It correlates requests, limits, actual usage, and node costs into a single view — giving every team full Kubernetes cost visibility, from namespace to workload, so they can see exactly what they're consuming and where the waste is.on cost control and optimization.

Works across EKS, AKS, and GKE

Whether you run Kubernetes on AWS, Azure, Google Cloud, or across multiple providers — OneLens gives you the same depth of visibility everywhere. One platform, one consistent experience, regardless of where your clusters run. No per-cloud workarounds, no fragmented tooling.
Amazon EKS
Google GKE
Azure AKS

Three capabilities, one platform

OneLens covers the full Kubernetes FinOps lifecycle across three integrated capabilities
Cost Visibility

Understand exactly what you're spending, at every level of your cluster hierarchy, with real utilization and efficiency metrics.

Optimization

Get safe, data-backed rightsizing recommendations based on 30 days of actual usage — not snapshots or estimates.

Cost Allocation

Attribute costs accurately to teams, products, or environments using namespace-based or label-based allocation, including idle and shared costs.

Real Kubernetes Savings, Not Projections
Most teams over-provision clusters by 30-40% without realizing it. OneLens helps platform teams find and fix the waste — rightsizing nodes, eliminating zombie instances, and consolidating under-utilized clusters. Customers typically reduce Kubernetes costs by 20-30% within weeks.
25%
Average Cost Reduction
14 Days
Time to First Savings
20+
K8 Cost Savings Policies

$26K Saved in 30 Days: No More Ghost Nodes

OneLens automatically detected 4 orphaned EKS nodes running at less than 1% CPU — completely invisible to our team. We terminated them within a sprint and saved over $2,200/month. That's $26,000 a year we were just burning. The platform didn't just flag the issue, it gave us the evidence, the ticket, and the recommended action. Zero manual investigation needed

VP of Engineering, Enterprise FinTech Platform

FAQ

What makes OneLens different from Cloud Native tools?

OneLens brings costs of all your cloud costs across multi-cloud(AWS, Azure, GCP, OCI), Kubernetes (EKS, AKS, GKE), AI, and your custom sourcesat one place.  OneLens allows you to allocatecosts and set budgets across all these sources to your businessconstructs such as Business Units, Applications, Projects, Environments,etc.  Then, you can setup customworkflows which deliver precise, timely, and relevant cost alerts andanomalies to teams and individuals who are accountable. No Cloud Native toolgives you deep, unified Kubernetes cost visibility – across requests,limits, usage, and nodes  at on place:OneLens does.

What is unique to OneLens?

Our goal is to bring all your Cloud, AI, and Kubernetescosts at one place; make it super easy for respective teams to understand andoptimize costs.Our Kubernetes Cost Visibility, Workflow Engine, and DeepInsights (storage, networking, data transfer) module stands out, solving thehardest and most time consuming problems of engineering and devops teams.

How quickly can we start seeing value? 

Most customers see their first cost and optimizationinsights within 36 hours of connecting cloud accounts.

How does OneLens handle Kubernetes costs? 

OneLens integrates directly with your clusters to break downcosts by namespace, pod, deployment, service, and any custom labels you use. Wecalculate true cost including shared cluster resources (control plane, systempods), ingress/egress traffic, and persistent storage-then allocate everythingbased on actual resource consumption. For organizations with large Kubernetesspend, this transforms opaque infrastructure costs into team-levelaccountability.

What Kubernetes distributions does OneLens support? 

OneLens works with all major Kubernetes distributionsincluding Amazon EKS, Azure AKS, and Google GKE. Our integration is read-onlyand doesn't affect cluster performance or require changes to your workloads.

Do you deploy an agent? If yes, how much does it cost?

Native Cloud tools do not collect adequate data for you tounderstand Kubernetes costs. We deploy an agent to bring all all that data togenerate required insights. You can learn all about it here. The cost is minimal and all details are here.

Can OneLens track costs by namespace, pod, or deployment? 

Yes. OneLens provides granular cost visibility at everylevel—cluster, namespace, deployment, pod, and even individual container. Youcan view costs by any Kubernetes label including team, application,environment, or customer ID. This enables precise cost allocation andchargeback for shared clusters.

How does OneLens allocate shared Kubernetes costs? 

OneLens gives you precise per-pod CPU/memory cost, whichallows you to allocate costs by namespaces, workloads or labels as per yourbusiness requirements. Further, OneLens automatically distributes shared cluster costs (idle costs, network costs) across pods. This ensures fair costdistribution and budgeting.

Do we need to change how we run Kubernetes to use OneLens? 

No. OneLens works with your existing Kubernetes setup and doesn't require changes to your deployment patterns, namespaces, or label structure.

What cloud providers does OneLens support? 

OneLens natively supports AWS, Microsoft Azure, Google CloudPlatform and Oracle Cloud. We ingest billing data from all three providers andnormalize it into a unified view so you can manage multi-cloud costs in oneplatform without switching between native tools.

How does OneLens connect to our cloud accounts? 

We have read-only access carefully designed to make sure weonly have billing, resource and usage data without getting access to any ofyour PII/customer data.

Does OneLens support FOCUS format?

Yes, we do. While you can analyze costs for each Cloud, youcan also look at costs across cloud using FOCUS format. This allows you tocreate cost centers across cloud spends and setup budgets.

Can OneLens track AI workload costs? 

Yes. OneLens tracks costs for AI platform services like OpenAI, AWS Bedrock, Azure AI Foundry, Google Vertex AI across all cloud providers.We show you exactly what your AI experimentation and production workloads cost.We are adding more AI providers every week.

How does OneLens help control AI spending? 

OneLens enables you to set budgets for AI projects, trackspending against forecasts, and get alerted when costs exceed thresholds.Combined with cost allocation, this gives you visibility into which AIexperiments or models are driving costs—helping you make informed decisionsabout resource allocation and project prioritization.

How does OneLens pricing work? 

OneLens uses transparent usage-based pricing calculated asper your managed cloud spend (Excluding taxes!). We have tiers to ensure youcan grow during the subscription term and not be surprised by any hidden charges. You only pay based on the cloud infrastructure we help you optimizeand control.

Is there a minimum spend requirement? 

OneLens is built fororganizations managing significant cloud infrastructure—typically $50K+ monthlycloud

Can we try OneLens before committing? 

Yes. We offer proof-of-concept engagements where you connect OneLens to your actual cloud data and validate the value before signing anannual agreement. Most POCs run for two weeks and include full platform accesswith implementation support.

Do you offer flexible contract terms? 

We offer both annual agreements and flexible month-to-month arrangements on AWS Marketplace depending on your organization's procurement cycles and budget planning. Contact our team to discuss what works best foryour situation.

What certifications do you have?

We are SOC 2 Type 2 Compliant and ISO 27001 certified.  Please check out our trust center for moredetails.

What access does OneLens need to our cloud accounts? 

OneLens primarily requiresread-only access to your billing and usage data. Further, the read-only accessis carefully designed to ensure that we don’t get access to data in yourlogs/buckets.

How is our data secured? 

OneLens is built with enterprise security standardsincluding encryption at rest and in transit, SOC 2 compliance, role-based access controls, and regular security audits. Your cost data is isolated indedicated environments and never shared across customers.

Where is our data stored? 

OneLens stores datain secure, encrypted databases within your choice of AWS/GCP regions. We canaccommodate

Can we control who sees what data in OneLens? 

Yes. OneLens includes granular role-based access controls,allowing you to restrict visibility by any org dimension through construct ofcost centers. Finance teams can see enterprise-wide data while individual teamssee only their allocated costs.

What kind of support does Astuto.ai provide? 

We strongly believe that FinOps is not about softwaresubscription. We assign a named, dedicated Cloud FinOps expert as your customersuccess and support point of contact. We work closely with finops, engineeringand business teams to ensure they are on top of costs, leaks and savings.

How do we get help if we have questions? 

You can connect with our team directly on Slack, Teams andEmail.  We have detailed documentation at https://docs.onelens.cloud

What does initial onboarding look like? 

Initial cloud onboarding is quick – just takes 10-15 minutesto deploy our automated IaC scripts to give us access. Post that it will take 24-48 hours for the initial data ingestion before we can share insights on yourcloud infrastructure.  Kubernetes onboarding takes a bit longer since it involves deploying agents, and waiting for enough data to generate insights.