Delivery Matrix
| Channel | Description | Status | Notes |
|---|---|---|---|
| API | REST access per dataset ID (/api-reference) | ✅ Supported | Ideal for filtered pulls and low-latency queries. See API intro. |
| SFTP | Managed SFTP drops with nightly dataset snapshots | ✅ Supported | Includes archived data and schema manifests. See SFTP quickstart. |
| WSS | Real-time event stream (<150 ms latency) | 🟡 In Development | Provides live updates for subscribers. See WSS reference. |
| MCP | Machine-to-Machine protocol for AI Agents | 🟡 In Development | Enables autonomous agent data retrieval and reasoning. |
| Amazon Redshift | Data sharing via Redshift Serverless | ✅ Supported | See Redshift Share. |
| Azure Blob Storage | ADLS Gen2 integration via service principal | ✅ Supported | See Azure Blob. |
| Snowflake Share | Secure Data Share to your Snowflake account | ✅ Supported | See Snowflake Share. |
| Google BigQuery | Authorized view or transfer service | ✅ Supported | See BigQuery. |
| Amazon S3 | Cross-region replication into customer-owned buckets | ✅ Supported | See S3. |
| Databricks Delta Share | Native Delta Sharing tables | ✅ Supported | See Delta Share. |
Most customers combine the API (for ad-hoc lookups) with one or more bulk delivery channels for analytics and archival needs.
Use the
Use the
_tracing_id column to reconcile records consistently across delivery channels.Choosing a channel
- Real-time pipelines → API / WSS
- Historical backfills → SFTP / S3 / Azure Blob
- Analytics at scale → Snowflake / Redshift / BigQuery / Delta Share