Sequin captures real-time changes from Postgres databases, ensuring that no insert, update, or delete is missed while providing exactly-once processing of all data modifications. This technology enables developers to implement asynchronous triggers and streaming capabilities without the operational complexity associated with traditional message queuing systems like SQS or Kafka.
Funding
Funding not disclosed
Founders
Product
Problem
Implementing asynchronous triggers and streaming capabilities for Postgres databases often involves operational complexity associated with traditional message queuing systems like SQS or Kafka. Ensuring exactly-once processing of data modifications without missing any inserts, updates, or deletes can be challenging and resource-intensive.
Solution
Sequin is a change data capture (CDC) platform for Postgres that captures real-time changes, ensuring no insert, update, or delete is missed, while providing exactly-once processing of all data modifications. It simplifies streaming Postgres data to various destinations, including streaming platforms, queues, and search indexes, without the operational overhead of traditional message queuing systems. Sequin uses a logical replication slot to guarantee no changes are missed and can capture old values for updates and deletes. It is delivered as a standalone Docker container, and a hosted offering is also available.
Target Audience
Sequin is designed for developers and organizations that require real-time data streaming from Postgres databases to various destinations for use cases such as triggering workflows, informing downstream services, audit logging, and maintaining caches and search indexes.
Features
- Real-time change data capture from Postgres databases (version 12+)
- Exactly-once processing of data modifications
- Logical replication slot to ensure no missed changes
- Capture of old values for updates and deletes
- Support for backfilling existing rows and streaming new changes in real-time
- Filtering and routing of messages to sinks using SQL `where` conditions or complex functions
- Message payload transformations with custom functions
- Infrastructure-as-code management using YAML files and a CLI, or a Management API
- Prometheus metrics endpoint for observability
- Web console for configuration and monitoring
- Support for various sinks, including Kafka, GCP Pub/Sub, HTTP endpoints, SQS, Kinesis, Elasticsearch, and Typesense