Resources

Q: How fresh is the inventory?

Default refresh is hourly. Critical attributes (state, tags) can be set to 5-minute polling on Pro/Enterprise.

Q: Can I export the inventory?

Yes — CSV, JSON Lines and a read-only API. Useful for CMDB sync.

Every VM, disk, bucket, function, queue, cluster and managed service across AWS, GCP, Azure, Oracle Cloud and Kubernetes — normalised, taggable and joined to the cost line that paid for it.

5 clouds + K8s
Cost per resource
Bulk tag editor

Start free

← All features

lumicost.com — resources

By the numbers

Resourcesin 4 numbers

AWS · GCP · Azure · OCI

clouds with live adapters

resources · metrics · cost

adapter roles per cloud

100% provider APIs

agents installed in your runtime

type · region · tags · metrics · cost

shared Resource model across clouds

How it works

Three steps. No agents. No surprises.

Connect read-only

Add an AWS account through a cross-account IAM role (read-only). GCP and OCI use the same pattern. We never gain write access to your accounts.

read-only IAM · STS:AssumeRole

Discover

For each connected account/project/subscription/tenancy, the matching adapter walks the configured regions and calls the cloud's resource API (AWS EC2, GCP Compute, Azure Resource Manager, OCI Resource Manager).

region-by-region crawl

Enrich & query

Resources are enriched with usage metrics (CloudWatch · Cloud Monitoring · Azure Monitor · OCI Monitoring) and monthly cost (Cost Explorer · Cloud Billing/BigQuery · Azure Cost Management · OCI Usage API). Query via /api/v1/resources.

metrics + cost · per cloud

What's inside

Every VM, disk, bucket, function, queue, cluster and managed service across AWS, GCP, Azure, Oracle Cloud and Kubernetes — normalised, taggable and joined to the cost line that paid for it.

One schema for five clouds

An EC2 instance, a GCE VM and an Azure VM are all 'compute' — and we model them that way. Filter, group and report without learning each provider's billing taxonomy.

Cost-per-resource, not just service

We join the line items in the cloud bill to the actual ARN / resourceId — so you can ask 'how much did THIS pod cost last week?' and get an answer in seconds.

Lifecycle signals built in

Each resource carries observed status: idle, orphaned, drifted (vs. IaC), about-to-renew, missing required tags. The signals also drive the recommendation engine.

Capabilities

Everything you need on day one.

AWS: EC2 · CloudWatch · Cost Explorer · EKS
GCP: Compute · Cloud Monitoring · Billing API + BigQuery · GKE
Azure: Resource Manager · Monitor · Cost Management · AKS
OCI: Compute / Block / Object · Monitoring · Usage API
Read-only credentials only
Multi-account · multi-region per cloud
Tag-aware filtering & search
One Resource model across providers

Before vs after Lumicost

Night and day.

Without Lumicost

Four cloud consoles, no shared model
Tags live in the cloud, cost lives elsewhere
Need an agent in every cluster/VM

With Lumicost

One Resource model across AWS · GCP · Azure · OCI
Tags + monthly cost on the same row
Provider APIs only — no install

Where teams use it

Three ways teams put it to work.

M&A discovery

Connect the acquired company's accounts and get a full inventory with owner attribution in under an hour.

Tag hygiene sprint

Find every resource missing a 'team' tag, group by service, fix in bulk via a CSV round-trip.

Cluster-wide K8s rightsizing

Filter to all pods with utilisation < 30% across all clusters, sort by cost, ship a single PR.

From the field

“If we have to install a daemon in your runtime, we've already lost. The Resource model is built only on what the cloud provider already exposes.”

The Lumicost teamdesign principle

Plays nice with your stack

AWSGCPAzureOracle CloudKubernetesTerraformPulumiSlackJiraLinearGitHubGitLabDatadogSplunkOpenTelemetryPagerDutyNetSuiteSAPAWSGCPAzureOracle CloudKubernetesTerraformPulumiSlackJiraLinearGitHubGitLabDatadogSplunkOpenTelemetryPagerDutyNetSuiteSAP

FAQ