Skip to main content

GCP Ai Platform Pipeline Job

A GCP AI Platform Pipeline Job (now part of Vertex AI Pipelines) represents a managed execution of a Kubeflow pipeline on Google Cloud. It orchestrates a series of container-based tasks—such as data preprocessing, model training, and deployment—into a reproducible workflow that runs on Google-managed infrastructure. Each job stores its metadata, intermediate artefacts and logs in Google-hosted services, and can be monitored, retried or version-controlled through the Vertex AI console or API. For full details, see the official documentation: Vertex AI Pipelines – Run pipeline jobs.

Supported Methods​

  • GET: Get a gcp-ai-platform-pipeline-job by its "name"
  • LIST: List all gcp-ai-platform-pipeline-job
  • SEARCH

gcp-cloud-kms-crypto-key​

A pipeline job can be configured to use customer-managed encryption keys (CMEK) so that all intermediate artefacts and metadata produced by the pipeline are encrypted with a specific Cloud KMS crypto key. Overmind therefore surfaces a link to the gcp-cloud-kms-crypto-key that protects the job’s resources.

gcp-compute-network​

Pipeline components often run on GKE clusters or custom training/serving services that are attached to a VPC network. When a job specifies a network or privateClusterConfig, Overmind links the job to the corresponding gcp-compute-network, highlighting network-level exposure or egress restrictions that may affect the pipeline.

gcp-iam-service-account​

Every pipeline job executes under a service account whose IAM permissions determine which Google Cloud resources the job can access (e.g. storage buckets, BigQuery datasets). Overmind connects the job to that gcp-iam-service-account so that permission scopes and potential privilege escalations can be inspected.

gcp-storage-bucket​

Pipeline jobs read from and write to Cloud Storage for dataset ingestion, model artefact output and pipeline metadata storage. Any bucket referenced in the job’s pipeline_root, component arguments or logging configuration is linked here, allowing visibility into data residency, ACLs and lifecycle policies relevant to the pipeline’s operation.