Core concepts

The five objects that make up a Chartizer team, and how they depend on each other.

Everything in Chartizer is built from five objects. Learn these and the rest of the product is mostly navigation.

Source

A source is one connected system plus the credentials and runtime used to reach it.

Created once, reused by every dataset drawn from it.
Counts against your plan limit — the Overview tab shows 8 / 20 Active / Limit.
Has a runtime: cloud (Chartizer connects) or agent (a local agent connects on your behalf).
Has a health state. When credentials expire the source drops into Needs attention with Authentication expired and every dataset beneath it stops syncing.

Sources hold credentials, datasets hold shape

Rotating a database password is a change to the source. Renaming a column is a change to the dataset. Keeping these separate is why one re-auth can fix a dozen broken datasets at once.

Dataset

A dataset is a named, typed view over part of a source — one table, view, index, query result, or uploaded file. Projects read datasets, never sources directly.

Each dataset carries a type and a status:

Status	Meaning	What to do
`Ready`	Last sync succeeded, schema is stable	Nothing
`Syncing`	A refresh is running now	Wait; the row updates itself
`Auth error`	The parent source cannot authenticate	Reconnect the source
`Schema drift`	Column types changed since last sync	Review and accept the new schema

Mapping

Coming soon

Mappings are still in development — see Mappings. Until they ship, a chart or an investigation works within a single dataset.

A mapping will reconcile fields across datasets so they can be joined and charted together: customer_email, Email Address, and contact.email becoming a single logical field.

Team, folder, project

A team is the top level — a self-contained Chartizer environment with its own members, roles, permissions, AI settings, billing, and data catalogue. Inside it, work is organised in two levels.

Level	What it is	Holds
Team	The environment itself: people, permissions, data, billing	Folders + the catalogue
Folder	A way to group related work	Projects
Project	The unit of work and of access	Charts, dashboards, files

Text

Team
├── Data catalogue          global to the team
│     sources · datasets · mappings
└── Folders
    ├── Internal/                   ← folder
    │   ├── Weekly revenue              ← project
    │   └── Dealer performance
    ├── Demos/
    │   └── Customer walkthrough
    └── Auto dīleriem POC/
        └── Campaign attribution

The catalogue is team-wide; access to it is not

Sources, datasets, and mappings belong to the team, so every project draws from one catalogue and you connect a database once. That is a convenience, not a permission: opening a dataset still requires it to be added to a project you are a member of.

Projects remain the unit of access control: granting someone a project grants that project, not its folder, not the team, and not the sources behind its datasets.

Agent

An agent is a small process you run inside your own network. It holds database secrets locally and executes queries there, returning only results to Chartizer.

Use an agent when:

The database is not reachable from the public internet.
Credentials may not leave your infrastructure for policy reasons.
You want query execution to stay on your hardware.

See Local agents for setup.

How they depend on each other

Text

Team
│
├── Agent (optional)
│     └── Source ─────────► credentials + runtime
│           └── Dataset ──► shape + sync status
│                 └── Mapping ► field reconciliation
│
└── Team
      └── Folder
            └── Project ──► charts, dashboards, files
                              └── reads Datasets ↑

Datasets sit on the team side of that split, which is why one source can feed projects in several teams at once.

Deleting an object always warns about what depends on it: removing a source removes its datasets, which breaks charts in every project reading them — including projects in teams you are not a member of. The confirmation names them.