Content Engine

A Content Engine is a self-contained setup for generating and managing content. It owns your data, the rules for writing it, the workflow that runs generation, and the finished content — all isolated from every other engine.

Engine = pipeline in the API

The studio calls this a Content Engine. In the Data API and URLs the same thing is still called a pipeline — they're interchangeable.

An engine has four parts:

Data

Your structured data, stored as objects in one or more data pools. Learn more

Workflow

The visual node graph that defines what happens to your data — transform, generate, quality-check, route, and deliver. This is where preprocessing and fanout behaviour now lives; there are no separate scripts. Learn more

Intent

What axite should write: Declarations (Stochastic AI) and Rulesets (Deterministic NLG). Learn more

Content Assets

The generated content. One object can produce many; all are kept with history. Learn more

You get data in and content out through the Data API and webhooks.

How generation runs

When generation runs — manually, or automatically when an object changes:

1. The Workflow executes for each object, starting at its Entry Point.
2. Generation nodes write content from your Intent; routing nodes split per language, segment, or variant.
3. Guardrails score each result, and a gate passes, holds, or rejects it.
4. Deliver nodes send the finished content to your systems via webhook.

You can monitor runs and results in the engine throughout.