Extending Looker "Self-Service Explores" (File Upload to Explore) to OEM/Embedded Users

Looker Q&A Looker
, , ,
1

Current Setup:

  • End-users interact with analytics strictly via Signed SSO Embedded Iframes within our host application.

  • End-users do not have direct access to the Looker instance or admin panel.

Requirement:

  • We want to expose the Self-service Explores experience (uploading ad-hoc CSV/Excel files or querying ad-hoc BigQuery tables) directly to our external embedded end-users inside our host app UI.

Key Questions for Google Engineering:

  1. What is the recommended pattern or API workflow to expose file-upload ad-hoc exploration (upload_data workflow) to SSO-embedded users without exposing the core Looker UI?

  2. Can we programmatically trigger the underlying Looker dataset generation pipeline (file \\rightarrow PDT schema generation) via the Looker REST API or Looker SDK?

  3. Are there specific database write permission or licensing considerations when embedding ad-hoc user data ingestion in an OEM SaaS architecture?

2

Based on the current API documentation, Self-service Explores can be configured for user uploads, but there is no public API for initiating the actual CSV/XLS upload and generated-model workflow. The available SelfService API methods mainly expose allowed connections, generated LookML, and certification controls. For an OEM setup, the practical pattern is therefore to provide the supported Looker upload UI through signed embedding, with appropriate permissions and connection governance, or build your own ingestion pipeline into BigQuery and expose the resulting tables through managed LookML. Database write access, tenant isolation, storage lifecycle, and Looker licensing should be reviewed with your Google account team before exposing this to external users.