Connectors & CDC

Every source, one typed ingestion layer.

Point SpooqW at your databases, streams and files. Each connector is a typed, validated step — the engine knows the schema before the run starts, and credentials never touch a log line.

Databases via JDBC

Postgres first-class, plus any JDBC database — partitioned parallel reads, predicate pushdown and typed column mapping out of the box.

Kafka ingestion, drained on a schedule

Each scheduled run consumes exactly what has arrived since the last one and lands it as governed Iceberg tables. JSON and Avro decode inside the run against the Avro schema the step carries — bring your own schema, there is no registry integration yet. Not a always-on streaming job: a bounded batch you can reason about, retry and receipt.

CDC through Debezium topics

Typed connector templates for the Debezium topics of Postgres, MySQL and SQL Server, plus DB2's read-log. The change events land as a typed input; a SQL step unpacks the envelope and an Iceberg MERGE writes the upserts and deletes, so silver tables stay in sync without full reloads. Debezium itself stays yours to run.

Object storage & files

CSV, JSON, Parquet and Avro from S3-compatible object storage, with schema inference validated into typed steps you can trust in production.

Hosted Postgres database

No source yet? On the cloud tiers, provision a managed Postgres in seconds — no VPN, no IP whitelist. You get a connection string and load data with psql, dbt or your own app; it's registered as a pipeline source automatically. Need new credentials? Delete and re-provision — the old database and role are dropped immediately, nothing is silently rotated behind your back.

Encrypted, masked credentials

Connection secrets are encrypted at rest with AES-256-GCM and masked everywhere they could appear — API responses, logs, audit records.

A growing catalog

The in-product catalog separates end-to-end verified connectors from read-only compatibility evidence. Relational databases, Kafka, object storage, MQTT, mailboxes and Iceberg are available only when their live proof and runtime capability are present in the deployed build. No packaged SAP or Salesforce connector is claimed: use a verified database or export path instead.