GCS Connectors

Bulk-import documents from Google Cloud Storage buckets into the ingestion pipeline via POST /v1/connectors/gcs/import.

GCS Bulk Import

POST /v1/connectors/gcs/import

Bucket name without gs:// prefix
Only import objects with this prefix

How GCS Import Works

1
Object Enumeration
The ingestion service lists all objects in the bucket matching the optional prefix filter.
2
Pub/Sub Queue
Each object is queued as an individual ingest job in Pub/Sub for parallel processing.
3
Document Processing
The processor chunks each document, generates embeddings via Vertex AI when configured (or local deterministic embeddings by default), and stores results in PostgreSQL with optional Vertex Vector Search indexing.
✓
Ready for Query
Documents with job status SUCCEEDED are indexed and available for RAG queries.

Import History

This session

BucketPrefixTenantJobs QueuedTime
No imports yet