Google Cloud Platform Integration

PromptQL for Google Cloud Platform

Connect Google Cloud Platform to PromptQL to get AI-powered cloud monitoring and billing analysis. Surface cost spikes, inspect cloud resources, investigate incidents, and connect infrastructure activity to the rest of your business.

Getting Started

How to Connect Google Cloud Platform

Once connected, Google Cloud Platform becomes a fully integrated infrastructure workspace in PromptQL. Analyze billing trends, inspect resource usage, review IAM and inventory, and connect cloud activity to incidents, deploys, and business signals using natural language.

Instructions to connect Google Cloud Platform

  1. 1

    Click on Data in the left pane.

  2. 2

    Search for Google Cloud Platform in integrations.

  3. 3

    Click Connect.

  4. 4

    Complete the OAuth flow.

  5. 5

    Done!

What you can do with PromptQL for Google Cloud Platform

AI-powered infrastructure analysis that turns cloud telemetry into action

What changed in our GCP spend this month and which services are responsible?

PromptQL

Analyzed billing across 12 projects. Total spend is up 21% month over month, driven mostly by BigQuery query costs (+38%), GKE node hours (+19%), and Cloud Storage egress (+27%). The sharpest spike began three days after the new analytics pipeline rollout in the data-prod project.

Break down spend by projectWhat changed after the rollout?

Which compute resources in GCP look underutilized or safe to rightsize?

PromptQL

Reviewed 146 compute resources across GCE and GKE. 19 VM instances average under 12% CPU utilization, 7 node pools show sustained overprovisioning, and 11 persistent disks are unattached. Estimated monthly savings from rightsizing and cleanup is $18.4K.

Show the low-utilization resourcesEstimate savings by project

Investigate yesterday's incident and show me which GCP services were affected first

PromptQL

Correlated logs and metrics across Cloud Run, Pub/Sub, and Cloud SQL. Earliest signal was elevated Pub/Sub delivery latency at 09:13, followed by Cloud Run 5xx spikes at 09:17 and Cloud SQL connection saturation at 09:21. Impact peaked in the customer-events project and subsided after the autoscaling change at 09:34.

Show the timelineWhich services recovered last?

Audit IAM and tell me which accounts have overly broad access in production

PromptQL

Reviewed IAM policies across 8 production projects. Found 13 principals with editor-level access and 4 service accounts with owner permissions that appear broader than necessary. 2 former employee accounts still exist in inherited bindings, and 6 external identities can read production storage buckets.

List the risky principalsRecommend least-privilege changes

Summarize our current cloud footprint and highlight anything unusual in inventory

PromptQL

Inventory scan found 812 total resources across prod, staging, and analytics projects. Unusual findings: 3 orphaned load balancers, 2 unused reserved IPs, 1 unexpectedly large Memorystore instance in staging, and 4 GCS buckets with public access prevention disabled. The biggest growth area this quarter is Cloud Run service count, up 42%.

Show the public bucketsCompare prod vs staging footprint

Unlock the full power of PromptQL

Replace siloed dashboards with cross-app intelligence that connects cloud operations to engineering, support, and finance workflows.

Correlate our GCP incident timeline with recent GitHub deploys and Slack escalation threads

Referenced:githubslack
PromptQL

Matched 6 production deploys, 3 GitHub workflow runs, and 11 Slack escalation threads to the incident window. The strongest correlation is a config rollout merged at 08:58, deployed at 09:05, and first mentioned in Slack at 09:16. Two later deploys appear unrelated and happened after recovery began.

Show the deploy linked to the incidentWhich Slack thread had the first report?

Compare GCP infrastructure spend with Stripe revenue and tell me where margins are compressing

Referenced:stripeslack
PromptQL

Compared 90 days of GCP spend against Stripe revenue by product line. Gross margin compression is most visible in analytics workloads, where infrastructure costs rose 17% while related revenue grew only 6%. The enterprise ingestion feature remains healthy, with revenue up 22% against a 9% cost increase.

Break down by product lineWhich workloads are driving the cost rise?

Show support-impacting GCP issues and the Zendesk tickets they are likely tied to

Referenced:zendeskgithub
PromptQL

Linked 4 GCP service degradations to 37 Zendesk tickets opened within the same windows. The most common customer-facing symptoms are delayed exports, timeouts on dashboard loads, and webhook retries. 12 tickets appear connected to a single Cloud SQL saturation event affecting enterprise reporting.

Show the related ticketsWhich issue affected the most customers?

Start querying your cloud infrastructure with AI

Connect Google Cloud Platform to PromptQL and get billing analysis, resource monitoring, incident intelligence, and cross-app infrastructure context - all in natural language.

No credit card required