Skip to main content
Data MartianOur data engine is a real-time platform built on the Entity-Component-System (ECS) pattern borrowed from game engines. Every piece of data is an entity, and all changes propagate in real time to connected users and background services.

​
Why ECS?

Traditional healthcare systems lock data into rigid table schemas. Adding a new field requires migrations, downtime, and risk. ECS inverts this:
  • Entities are generic containers — they don’t care what data they hold
  • Components define the data — add or remove them freely without schema changes
  • Systems react to data — subscribe to what they care about, ignore the rest
This means you can attach an Insurance component to a Claim entity today, and a DenialReason component tomorrow, without touching the database schema.

​
Core layers

The data engine is organized into three layers:
  1. API Layer — REST API for reads/writes, WebSocket for real-time subscriptions
  2. Commit Layer — validates authorization, merges deltas, creates immutable checkpoints
  3. Distribution Layer — fans out changes to users (via pub/sub) and services (via durable message queue)

​
Key properties

Real-time

Every entity change is delivered to connected clients within milliseconds via WebSocket.

Immutable audit trail

Every change creates an immutable checkpoint recording the delta, previous state, and new state. Nothing is ever lost.

Composable data

Entities gain meaning from the components attached to them. The same entity framework represents claims, patients, accounts, and EDI files.

Query subscriptions

Clients subscribe to entities matching a query. When matching entities change, clients are notified. When entities stop matching, clients are told to refresh.