AI
How the AI behind DataMind Platform works: Minds, builders, helper agents and how personal data is handled.
- How a Mind answers
The response modes a Mind offers, how it clarifies, plans and paginates an answer, and how it behaves when tools are off.
- Charts, KPI cards and diagrams in answers
The blocks a Mind can add to an answer: charts, KPI cards, diagrams, and the references that link to queries and documents.
- Personal data in AI work
How personal data is handled while Minds and builders work with it: masked values, what the AI can do with them, and where requests are blocked.
- Model AI
Ask Model AI to read, check and change a semantic model: acceptance modes, the measure workflow, reverts, RLS limits and turn caps.
- How measures run
The execution rules behind a measure: joins, PARTITION BY ALL, QUALIFY, totals, helper measures and patterns to avoid.
- The Slides AI
What the Slides AI builds and changes while you have a presentation open: layouts, theme tokens, page numbers, motion and its checks.
- The Warehouse Builder
How the Warehouse Builder turns a request into tables and transformations: intake, build order, patterns, tests, triage and retries.
- Links, KPIs and diagrams in the Warehouse Builder
How the Warehouse Builder links the objects and people it names, and how to read the KPI cards, charts and diagrams in its answers.
- Build an ANY API connector
Work with the AI panel to build, test and confirm an ANY API connector, from the manifest to stream discovery and sampling.
- Connector hosts, secrets and data access
Approve the hosts an ANY API connector may call, enter its secrets in the connection form, and grant it access to read table rows.
- AI helper agents
The small AI helpers behind DataMind OS actions: column comments, key suggestions, transformation drafts, static-table fills and automatic names.
- AI in the SQL editor
The inline AI suggestions in the SQL editor and the query writer behind them: what they propose and when they stay quiet.