Estuary is a managed ETL and data integration platform for data teams running CDC, batch and streaming pipelines into warehouses, lakes and Kafka.
Estuary moves data out of operational databases, SaaS applications and event streams into the warehouses, lakes and topics where teams analyse it. It reads database transaction logs through change data capture, so inserts, updates and deletes arrive as they happen instead of in a nightly job. Each destination can receive data continuously or on a schedule the team sets. Estuary Technologies, Inc. builds and runs the platform; the company was founded in 2019 and is based in New York. Captures pull from sources such as PostgreSQL, MySQL and MongoDB and write into collections, which hold the data as it flows. Materializations deliver those collections to Snowflake, BigQuery, Databricks, ClickHouse or Apache Iceberg tables on object storage. Derivations transform data in flight with SQL or TypeScript, and redaction blocks or hashes a field before it is written. Schema evolution carries source changes through to the destination, and a backfill reloads one collection without rebuilding the rest. Dekaf, a Kafka-compatible API, lets existing Kafka consumers read collections directly. Data engineering teams that keep dashboards, operational applications and AI systems fed from production databases are the core users. It fits teams that want streaming and batch delivery in one tool rather than two. Engineers can build pipelines in the Estuary web app or define them as YAML specs and publish them with the flowctl CLI. Regulated organisations can run a private data plane, or one inside their own cloud account, while Estuary hosts the control plane. A typical job replicates a PostgreSQL or MySQL database into Snowflake or BigQuery so analysts query current rows rather than yesterday's export. Another lands change streams as Iceberg tables for a lakehouse, then triggers a dbt Cloud job when new data arrives. Real-time analytics teams send streams into ClickHouse, or into Tinybird and StarTree, two analytics services Estuary partners with. Teams that work through a coding agent can use Estuary's Agent Skills to draft a pipeline from a plain-language prompt. Task logs, data stats and email alerts in the web app show when a pipeline stalls, before a report goes stale.
Pricing
View plans-
Developer
Free
- Up to 10 GB of data a month and 2 connector instances
- No credit card required
- The full platform, with real-time or batch pipelines
- Exactly-once delivery, automatic schema evolution and in-flight transforms
- Custom connector development not included
-
Cloud
$100 per month
Billed monthly as you go, stop anytime; $100 a month per connector instance ($50 each after the sixth) plus $0.50 per GB of data
- 30-day free trial once usage passes the Developer level
- 100 GB a month on 2 connector instances comes to $250 a month
- At these rates, 1,000 GB a month on 2 instances comes to $700
- Everything in Developer, with no data volume cap
- Unlimited connector instances, each billed
- Custom connector development by Estuary on request
-
Enterprise
Custom pricing
- Price on request, annual contract with discounts
- Everything in Cloud
- Private deployment and Bring Your Own Cloud (BYOC)
- SSO, SOC 2 report and HIPAA BAA
- Dedicated support, custom SLAs and migration help
- Custom connector development included
Features
- Included: Managed source connector library
- Not included: Custom connector builder
- Included: Incremental loads after the first sync
- Included: CDC mode for database sources
- Included: Automatic schema drift handling
- Included: Sync frequency down to minutes
- Included: Column selection and exclusion
- Included: In-flight column hashing
- Included: Post-load transformations in the warehouse
- Included: Data lake destinations in open formats
- Included: Per-sync run logs and row counts
- Included: Sync failure notifications
- Included: Single-table resync
- Included: Pipelines managed by API or Terraform
- Included: Self-hosted data plane option
- Not included: Open source ETL engine
Additional Features
- Change data capture from database transaction logs
- Captures, collections and materializations
- Streaming or scheduled batch delivery per destination
- Exactly-once delivery
- Automatic schema evolution
- SQL and TypeScript derivations
- Field redaction and hashing at capture
- Per-collection backfills
Best for
- Data engineering teams replicating production databases into a warehouse
- Teams that want streaming and batch delivery from one tool
- Teams feeding Kafka consumers without running Kafka
- Regulated teams that need the data plane in their own cloud
- Analytics teams loading Iceberg tables for a lakehouse
- Developers who manage pipelines from a CLI
- AI teams keeping retrieval data in step with source systems
Use cases
- Replicating PostgreSQL into Snowflake with CDC
- Streaming MySQL changes into BigQuery
- Syncing MongoDB collections to a warehouse
- Landing change data as Apache Iceberg tables
- Feeding ClickHouse for real-time analytics
- Serving collections to Kafka consumers through Dekaf
- Triggering dbt Cloud jobs when new data lands
- Hashing personal fields before they reach a destination
- Backfilling one collection after a source change
- Keeping data for AI and RAG applications current
Screenshots & Videos
Explore Estuary in action
User Reviews
Compare with Similar Software
Click to compare side by side with Estuary