Connectors
50+ connectors. More every month.
Queryvine ships native connectors for the sources analytics teams actually run — with schema-drift detection built into every one. Connectors monitor INFORMATION_SCHEMA for warehouses, Avro/Protobuf schema registries for event streams, and REST API response structures for SaaS sources.
All connectors
Snowflake
Warehouse
Schema fingerprinting via
INFORMATION_SCHEMA.COLUMNS. Supports VARIANT column tracking, auto_migrate for additive changes, and MERGE-based upsert writes.BigQuery
Warehouse
Schema monitoring via BigQuery INFORMATION_SCHEMA. Supports streaming inserts and MERGE DML. Detects STRUCT field changes in nested columns.
Redshift
Warehouse
Monitors
pg_catalog.pg_attribute for column-level changes. Uses COPY-based bulk loads with staging table swap for idempotent writes.Databricks
Warehouse
Schema monitoring via Unity Catalog
INFORMATION_SCHEMA.COLUMNS. Tracks Delta Lake column evolution events including ALTER TABLE ADD COLUMN and type promotions.Apache Kafka
Streaming
Schema drift detection at the consumer group level. Monitors Avro and Protobuf schemas via Confluent Schema Registry. Intercepts incompatible schema evolution before records are deserialized.
AWS Kinesis
Streaming
Samples shard iterators to infer payload structure via JSON schema inference. Detects field additions, removals, and type coercions across rolling 5-minute windows.
Pub/Sub
Streaming
Monitors Schema Registry entries via the Pub/Sub Admin API. Fingerprints Avro schemas per topic and alerts on backward-incompatible revisions before subscriber processing.
Confluent Cloud
Streaming
Polls Confluent Schema Registry REST API for schema version diffs. Supports FULL and FORWARD compatibility modes; blocks consumer lag spikes caused by unexpected schema breaks.
Postgres
Database
Monitors
pg_catalog.pg_attribute and pg_type for column-level drift. Uses logical replication slots to capture DDL events including ALTER TABLE operations in real time.MySQL
Database
Queries
information_schema.COLUMNS on each poll cycle. Parses binlog DDL statements via MySQL replication protocol to detect ALTER TABLE events between poll intervals.MongoDB
Database
Samples document field paths via Change Streams and infers collection schema using JSON Schema validation. Detects field-type coercions and unexpected
null field introductions.SQL Server
Database
Monitors
sys.columns and sys.types for DDL changes. Leverages SQL Server Change Data Capture metadata to track column additions and data type alterations per table.Salesforce
CRM
Polls Salesforce Metadata API for object field-level changes. Detects custom field additions, picklist value changes, and field-type reclassifications that break downstream pipeline expectations.
HubSpot
CRM
Uses HubSpot Properties API to fingerprint CRM object schemas. Monitors property group changes, custom property additions, and field-type mutations across Contacts, Deals, and Companies.
Stripe
Payments
Monitors Stripe webhook event payloads for structural drift. Fingerprints
data.object field sets per event type and alerts when Stripe API version upgrades introduce breaking payload changes.dbt Cloud
Transformation
Trigger dbt Cloud jobs after pipeline runs complete. Drift-aware execution gates: if a critical schema drift event occurred upstream, dbt jobs are held until the drift is resolved.
Fivetran SDK
EL-compatible
Wraps Fivetran HVR connector output with Queryvine schema fingerprinting. Intercepts
OpType metadata and ALTER_TABLE change events before records land in the destination.Airbyte SDK
EL-compatible
Integrates with Airbyte's
AirbyteCatalog JSON Schema manifests. Diffs catalog snapshots across sync runs to surface field additions, removals, and type array changes per stream.Amazon S3
Object Storage
Scans S3 prefix manifests for Parquet and JSON files. Infers and diffs column schemas across landing zone partitions; detects silent column drops from upstream export jobs.
GCS
Object Storage
Monitors GCS bucket prefixes via Storage Notifications (Pub/Sub). Reads Parquet/Avro footers to fingerprint file-level schemas and alerts when column structure diverges from the registered baseline.
Azure Blob
Object Storage
Subscribes to Azure Event Grid blob-created events. Reads Parquet metadata footers via Azure Data Lake Storage Gen2 API to track schema evolution within ADLS container paths.
SFTP / CSV
File
Polls SFTP drop zones for new CSV deliveries. Compares header rows and inferred column types against the registered baseline to detect vendor-side schema changes before ingestion.
Connector SDK
Build your own connector in 30 minutes
Queryvine's open connector SDK lets you add any source — internal databases, proprietary APIs, custom data formats. The SDK handles schema fingerprinting; you implement fetch_schema() and fetch_batch().
internal_metrics_connector.py
from queryvine.sdk import BaseConnector, Schema, Batch
class InternalMetricsConnector(BaseConnector):
"""Pull from internal observability DB."""
connector_id = "internal_metrics"
version = "1.2.0"
def fetch_schema(self) -> Schema:
# Queryvine calls this on each poll cycle
rows = self.db.execute("SELECT * FROM metrics LIMIT 0")
return Schema.from_cursor(rows.cursor)
def fetch_batch(self, since, until) -> Batch:
# Queryvine calls this to fetch data window
rows = self.db.execute(
"SELECT * FROM metrics WHERE ts BETWEEN %s AND %s",
(since, until)
)
return Batch(rows, schema=self.last_schema)
def validate(self, config: dict) -> bool:
return "connection_string" in configYour stack is already supported.
Connect a source in minutes. Schema drift protection included.