Workspaces & organizations

Workspaces

A workspace is the container everything else lives in: signals, segments, models, pipelines, trainings, deployments, and goals all belong to exactly one workspace. Data never crosses workspace boundaries.

Nearly every API request carries an X-Workspace-Id header that says which workspace the request operates on:

curl "$API_BASE/v1/signal" \
  -H "Authorization: Bearer $TOKEN" \
  -H "X-Workspace-Id: $WORKSPACE_ID"

from predictai import PredictAI

client = PredictAI(token="pa_live_…", workspace_id="ws_…")

data = client.signals.get_signal()

import { PredictAI } from "@predictai/sdk";

const client = new PredictAI({ token: "pa_live_…", workspaceId: "ws_…" });

const data = await client.signals.getSignal();

Omit it and you'll get a 400; send a workspace you can't access and you'll get a 403. Find your workspace IDs with GET /v1/workspace.

Workspaces also pin your data to a region: each workspace lives in one region, and your API base URL follows it. See Base URL & versioning.

Organizations

An organization groups people and workspaces under one roof — shared billing, shared membership, role-based permissions. You don't need one to use predictAI (individual accounts own workspaces directly), but teams usually want one.

Members have one of four roles:

When you belong to an organization, resources you create in its workspaces belong to the organization, and permission checks use your role — for example, deploying a model requires the models.deploy permission that member and above carry.

Manage all of this through the Workspaces section of the API: workspaces, members, organizations, roles, and invitations.