DataMind OS moves data in one direction. Sources feed the Lake, the Lake feeds the Warehouse, and the Warehouse feeds queries, Minds, reports, presentations and Writeback. Read this page to see which area does which job and where each one lives in the sidebar.
Under Data pipeline on the Dashboard, four numbered cards show the order of work:
Once the data is loaded and shaped, use it in one of these areas:
| Area | What you do there | Start here |
|---|---|---|
| Editor | Query and explore your data with SQL. | Editor |
| Minds | Ask anything across your data. | Minds |
| DataPrism | Build reports and explore semantic models. | DataPrism |
| Slides | Build and present decks from your data. | Slides |
In Warehouse, the Add New screen offers these objects:
| Object | What it is | Read more |
|---|---|---|
| Dimension | Descriptive data, such as customers, products or dates, that facts join to. | Create a dimension or fact table |
| Fact | Measurable events, such as sales or orders. | Create a dimension or fact table |
| Transformation | The query that fills a dimension or fact table from Lake data. | Build, test and finalize a transformation |
| Dictionary | Lookup values, filled by SQL queries and refreshed on a schedule. | Dictionaries |
| Static | Reference data you type or upload yourself, with no live source. | Static tables |
| View | An object defined by a SQL query. | Views |
| Semantic source and Logical model | Governance objects. Every dimension and fact table names a logical model and a source system, and each source system belongs to a semantic source. | Semantic sources, source systems and logical models |
Observability shows platform health and performance. Lineage traces each table from source to finish, and it covers tables, Minds, DataPrism models and Slides decks. Use it to see what an object depends on (upstream) and what depends on it (downstream). Start with Observability and Lineage.
Several areas hand their SQL to the Editor, which opens it in a new console:
| From | Button |
|---|---|
| A query in a Mind's answer | Open in Query Editor |
| The SQL of a DataPrism report chart | Open in Query Editor |
| The query of a Warehouse transformation or view | Open in Editor |
| The query of a Writeback table | Open in Editor |
| A query in Query Management, in Observability | Open in Editor |
These buttons appear only if you have Execute Query or Add/Remove Consoles.
A Mind is an AI agent you chat with about your data. You pick its kind when you create it.
| Kind | What it does |
|---|---|
| Analytical Mind | Answers questions over the tables you link in Sources. |
| Warehouse Builder | Builds facts, dimensions, views and transformations for you. It sees every Lake and Warehouse table and is owned by the admin role. |
| Lake Builder | Imports source tables into the Lake and manages connectors, syncs and load settings for you. It sees every Lake and Warehouse table and is owned by the admin role. |
The Warehouse Builder and Lake Builder options appear in the Mind form only if your role lets you create them.
The sidebar on the left lists every area your role can open. It has four groups, separated by lines:
Find your way around explains how to use the sidebar and the account menu, including the small BETA and SOON badges on some items and More, where items move when your window is too short to show them all.
The sidebar spells the area WriteBack and the area Admin. The Dashboard cards call them Writeback and Administration.
Your administrator gives you one or more roles, and each role carries permissions. The permissions decide which sidebar items you see:
| Sidebar item | Shown when you have |
|---|---|
| Workflows | View Lake Workflow Objects, View Warehouse Workflow Objects or View WriteBack Workflow Objects |
| Minds | View Minds |
| Editor | Execute Query or Add/Remove Consoles |
| DataPrism | View DataPrism Reports or View DataPrism Semantic Models |
| Slides | View Decks or View Others Decks |
| Lake | View Lake Objects |
| Warehouse | View Warehouse Objects |
| WriteBack | View WriteBack Objects |
| Lineage | Any one of View Lake Objects, View Warehouse Objects or View WriteBack Objects |
| Admin | Any Administration permission, for example View Users or View Roles |
| Observability | Any Observability permission |
Dashboard, Universe, Documentation and BI are shown to every signed-in user. An area you cannot open is not shown at all. A group with no visible items also loses its separator line. If you type the address of an area you cannot use, you see an Access Denied page.
The Flow view of the Dashboard draws the same picture with your own figures. Your connected sources are on the left, the platform is in the middle with figures for Data Lake, Data Warehouse, Data Sources, AI Minds and User Activity, and the outputs Minds, Editor, Slides, DataPrism and Writeback are on the right.
It is one of the Dashboard's three views; The Dashboard covers the view switch. To open each box and read what the figures count, see The Flow view.
The Lake holds raw data as it arrives from your sources. The Warehouse holds clean, modelled tables that are built from Lake data. Load the Lake first, then shape the Warehouse.
No. The Database Explorer in the Editor lists Lake tables as well as Warehouse tables, for the role you have selected, so you can query Lake tables directly. A Writeback table can also read straight from a Lake table. See Run your first query.
Use an Analytical Mind to ask questions about the tables you link to it. Use the Warehouse Builder or the Lake Builder when you want the AI to build Warehouse objects or import sources for you. See Create a Mind.
In Workflows. Each workflow has a type of Lake, Warehouse or WriteBack, and each one has its own schedule. See Schedules and when a workflow starts.
Use Writeback. Add a destination, which can be a database, Google BigQuery or RabbitMQ, and then add the tables that send data to it from a Lake or Warehouse table. You can also connect a tool you already use, such as a SQL client, Excel or Power BI, from Universe. See Writeback and Integrations.
DataPrism builds interactive reports on semantic models, which readers can filter and explore. Slides builds presentations, and a chart, table or text can be filled from a live query. Use DataPrism for reports people explore, and Slides for decks you present or email.