Integrations

Works with Kafka, Kinesis, Flink, Pulsar, and more

Broker-layer connection using read-only consumer groups. No changes to your producers, your consumers, or your topic configuration. No SDK to install.

Stream Sources

Streamforge connects using a dedicated read-only consumer group. It never writes to your topics, never modifies your existing consumer groups, and never touches your ACLs or IAM policies beyond what you explicitly grant for read access.

Apache Kafka
Confluent Cloud
Amazon Kinesis
Apache Flink
Apache Pulsar
Redpanda
MSK (Amazon Managed Kafka)

Downstream Stores

Read-only monitoring to detect when schema drift or data loss at the stream layer has already propagated into your warehouse. Useful for cross-layer correlation: "the Kafka topic started drifting at 03:14 UTC, and these Snowflake rows are suspect from that timestamp onward." We do not write to your warehouse or execute queries that modify data.

Snowflake
Google BigQuery
Amazon Redshift
Databricks Delta Lake

Alert Destinations

One alert per incident, not one per failed message. Configurable dedup window from 1 minute to 4 hours prevents alert storms when a single bad producer batch generates thousands of violations. Incidents auto-resolve in PagerDuty and OpsGenie when Streamforge detects the self-heal completed.

PagerDuty
Slack
OpsGenie
Microsoft Teams
Email (SMTP)
Webhooks

Don't see your stack?

Email us at [email protected]. We add integrations based on what customers actually run, not a roadmap document.