GCP coverage
Analyze costs and resource usage across Compute Engine, GKE, Cloud SQL, BigQuery, Cloud Run, and Agent Platform.
Investigate GCP usage
- Cost visibility
- See every dollar across GCP services, compute, storage, databases, BigQuery, and networking.
- Waste detection
- Identify unused persistent disks and inefficient logging configurations.
- Remediation tracking
- Send findings to Slack, Jira, Teams, or ServiceNow and track each change through resolution.
GCP coverage details
| Service or resource | What PointFive analyzes |
|---|---|
| Cloud Storage configuration | Storage class, lifecycle, and configuration choices silently inflate GCS bills. PointFive analyzes tiering, lifecycle, and configuration at the bucket level, including usage metrics and operation patterns. |
| Idle Agent Platform resources | AI/ML spend compounds quickly across endpoints, notebooks, and training pipelines, often without anyone noticing. PointFive detects idle endpoints, underutilized notebooks, and oversized training jobs across Agent Platform. |
| Cloud SQL storage | GCP Recommender handles CPU/memory rightsizing but doesn't analyze storage provisioning vs. actual usage. Over-provisioned Cloud SQL storage is invisible to native tools. |
| Managed instance group sizing | Google's MIG machine type recommendations cover only single-zone groups that are not autoscaled or load balanced. PointFive analyzes MIG templates and recommends machine type changes at the template level, rightsizing across all instances in the group. |
| Cloud Logging configuration | Log routing and storage tier configurations that silently accumulate cost. PointFive analyzes Cloud Logging bucket architecture for cost efficiency, routing verbose logs to cheaper storage. |
| GKE workload sizing | Agentless Kubernetes optimization for Deployments, DaemonSets, StatefulSets, and CRD-managed workloads on GKE, no sidecar or agent deployed on your clusters. |
Service or resourceCloud Storage configuration
What PointFive analyzesStorage class, lifecycle, and configuration choices silently inflate GCS bills. PointFive analyzes tiering, lifecycle, and configuration at the bucket level, including usage metrics and operation patterns.Service or resourceIdle Agent Platform resources
What PointFive analyzesAI/ML spend compounds quickly across endpoints, notebooks, and training pipelines, often without anyone noticing. PointFive detects idle endpoints, underutilized notebooks, and oversized training jobs across Agent Platform.Service or resourceCloud SQL storage
What PointFive analyzesGCP Recommender handles CPU/memory rightsizing but doesn't analyze storage provisioning vs. actual usage. Over-provisioned Cloud SQL storage is invisible to native tools.Service or resourceManaged instance group sizing
What PointFive analyzesGoogle's MIG machine type recommendations cover only single-zone groups that are not autoscaled or load balanced. PointFive analyzes MIG templates and recommends machine type changes at the template level, rightsizing across all instances in the group.Service or resourceCloud Logging configuration
What PointFive analyzesLog routing and storage tier configurations that silently accumulate cost. PointFive analyzes Cloud Logging bucket architecture for cost efficiency, routing verbose logs to cheaper storage.Service or resourceGKE workload sizing
What PointFive analyzesAgentless Kubernetes optimization for Deployments, DaemonSets, StatefulSets, and CRD-managed workloads on GKE, no sidecar or agent deployed on your clusters.
What BigQuery waste does PointFive find that native recommendations miss?
BigQuery is often the fastest-growing, least-owned line on a GCP bill. PointFive maps spend from the organization down to the project, dataset, table, and job level.
- Unused tables, and the pipelines still writing to ones nobody reads
- Datasets on the wrong storage billing model, logical vs. physical
- Temporary and non-production tables quietly accumulating cost
- Storage anomalies traced back to a specific dataset or table
- Coming soon: time-travel and fail-safe optimization, slot, reservation, and commitment optimization
82% lower BigQuery storage costs in 5 days
One billing-model optimization across 435 datasets generated $330K+ in annualized savings, enough to cover the cost of PointFive. PointFive analyzes storage behavior over time to identify when a different billing model can reduce costs, rather than relying on a single point-in-time comparison.
- Lower storage costs
- 82%
- Annualized savings
- $330K+
- Days to first savings
- 5
GCP services in context
Review Kubernetes workloads on GKE, AI training on Agent Platform, and BigQuery analytics alongside the infrastructure they use. Coordinate Cloud Storage lifecycle policies with application usage and investigate Cloud Logging configuration.
Enterprise-grade security and controls.
Independent audits, clear privacy commitments, and controls that keep your organization in charge. Explore our certifications, security reports, and access safeguards.
Visit our Trust CenterFrequently asked questions
For discovery, PointFive uses a read-only service account with minimal IAM permissions scoped to your project. Discovery uses read-only access. Automated remediation uses customer-authorized permissions, and each customer chooses how changes are approved: per change, once for a trusted playbook, or not at all. Connect through guided onboarding.
PointFive detects waste across Compute Engine, GKE, Cloud SQL, BigQuery, Cloud Run, Cloud Storage, Cloud Run functions, Gemini Enterprise Agent Platform (formerly Vertex AI), and more.
Google Cloud Recommender provides resource-level recommendations such as VM and Cloud SQL rightsizing. PointFive adds DeepWaste investigation of architectural waste, like Cloud Logging bucket inefficiencies and cross-service optimization.
Yes. PointFive breaks down BigQuery compute and storage spend by project, dataset, table, and job. It finds unused tables and pipelines still writing to data nobody reads, mapped from the jobs that actually ran, and flags datasets better suited to logical or physical storage billing. Time-travel and fail-safe optimization, plus slot, reservation, and commitment optimization, are coming soon.
PointFive analyzes BigQuery at the dataset level and looks at storage behavior over time to identify where a different billing model can reduce cost. In one customer environment, a billing-model optimization across 435 datasets cut storage costs 82% within five days, generating more than $330K in annualized savings, enough to cover the cost of PointFive.
Discovery uses read-only access. Automated remediation uses customer-authorized permissions, and each customer chooses how changes are approved: per change, once for a trusted playbook, or not at all. Dedicated service account with minimal IAM roles. Data encrypted in transit and at rest. SOC 2 Type II. No agents to install or maintain.
Yuval Goldstein detects 4 Cloud Logging inefficiencies live on a real GCP environment, and shows exactly how to fix them. 45 min, instant access.
