Agents

Agents

How agents work in Galaxy

Agents let anyone ask questions of your data in plain language and get back answers, reports, and automated workflows — grounded in the objects, relationships, and graph you've modeled.
  1. 1.

    Chat — answers, not queries.

    Ask a question the way you'd ask a colleague — 'which accounts are at risk of churning this quarter?' The agent finds the right objects, writes the query, runs it, and replies with the answer and citations back to the source data.

  2. 2.

    Workbooks — living reports.

    Turn a line of questions into a workbook: a structured report of metrics, charts, and tables that re-runs on fresh data every time you open it.

  3. 3.

    Automations — work that runs itself.

    Chain steps into an automation that triggers on a schedule or a signal — so a recurring digest, a churn-risk alert, or a draft-and-route job happens without anyone clicking run.

  4. 4.

    Grounded in your context.

    Because agents sit on top of your modeled objects and graph, answers stay consistent with how your business actually defines a customer, an order, or a churn risk.