Supported integration surfaces
Commerce and catalog connectors
Shopify, WooCommerce, BigCommerce, CSV, and S3 connectors help you get catalog data into NSL without building a custom ingestion pipeline first.1
Open Integrations
Go to Console > Integrations.
2
Choose the provider
Select Shopify, WooCommerce, BigCommerce, CSV, or S3.
3
Enter scoped credentials
Use read-only or narrowly scoped credentials wherever possible. NSL stores only the connection configuration needed to read catalog data.
4
Verify sync state
Confirm that catalog records appear in the console and that metadata fields are available for contexts, rules, and pipelines.
Provider requirements
Twelve Labs enrichment
The Twelve Labs integration adds async multimodal understanding for eligible video catalog items. When enabled, NSL can:- create and manage one team-scoped Twelve Labs index
- read configured metadata keys that contain public video asset URLs
- run enrichment outside the recommendation serving path
- store analysis metadata, vectors, and relationship signals for reuse
- let recommendation pipelines opt in to or ignore Twelve Labs-derived signals
Identity and collaboration integrations
The integrations page also surfaces NSL Connect workflows:- Partner IDs register activation or collaboration partners and mint partner-scoped NSL IDs.
- Data collaborations create approved match spaces between two NSL teams and calculate aggregate overlap.
- LiveRamp configuration can be used where a team needs an external identity resolution provider alongside NSL’s first-party identity graph.
API actions
The Platform API exposes integration actions throughPOST /api/integrations.
Common actions include:
connect_providerfor Shopify, WooCommerce, BigCommerce, CSV, and S3disconnect_providerfor connected providersconfigure_twelve_labssync_twelve_labsdisable_twelve_labs- default connection status updates for simple integrations
deletefor removing a saved integration record
Verification checklist
After connecting an integration, check:- catalog records appear in Users & Items or item search
- expected metadata keys are available for filters and rules
- ingest or sync timestamps are recent
- pipelines that depend on integration-derived signals are enabled intentionally
- credentials are scoped to read only what NSL needs

