How It Works
Data Estuary's architecture is built around three simple concepts that work together to handle your data logistics.
Entities
The What
Your data objects—Orders, Customers, Logs, whatever makes sense for your domain. Each has a unique ID and type that's recognized across all clusters.
- •Globally unique identity
- •Typed fields with validation
- •Optional state machines
Pipelines
The How
Step-based workflows that move, transform, and act on entities. Your business logic, triggered by events, schedules, or manual calls.
- •Sequential steps
- •Query, transform, update, create
- •Run on any cluster
Clusters
The Where
Physical locations where entities live and pipelines run. Deploy in cloud, on-premise, or edge—Data Estuary handles the logistics.
- •Built-in AWS or BYOC
- •Automatic replication
- •Flexible topology
How They Work Together
The power of Data Estuary comes from how these three concepts interact. You define what your data is (entities), how it moves and transforms (pipelines), and where it lives (clusters). Data Estuary handles all the complexity of replication, consistency, and orchestration.
Example: Processing an Order
An Order entity is created with status "pending"
A pipeline automatically triggers, checks inventory, and updates the order status
The updated order replicates across clusters based on your topology rules
Another pipeline on a different cluster detects the update and creates a shipment
No custom APIs to build. No message queues to manage. No replication logic to write. You focus on business logic—Data Estuary handles the logistics.
Why This Architecture
Separation of Concerns
Data logistics (replication, consistency, availability) is separate from business logic. Change your deployment topology without touching your pipelines.
Deploy Anywhere
Clusters can run in cloud, on-premise, or at the edge. Process data where it makes sense for your cost, latency, and compliance requirements.
Loosely Coupled
Teams coordinate through entities, not APIs. Change one pipeline without affecting others. No central ESB creating bottlenecks.
Infrastructure as Code
Define your entire topology—entities, pipelines, clusters—as code. Version control your data logistics alongside your application code.
Want to See It in Action?
Explore real-world use cases showing how these concepts solve enterprise problems.
View Use Cases