Skip to main content
Version: 0.0.42

Guides

This section contains task-focused step-by-step guides for all major features of Lakehousecat. For conceptual background on the semantic layer, modeling patterns, access control, and sharing, see Concepts.

Setup Workflow​

For a new Lakehousecat environment, the recommended setup order is:

  1. Configure Provider Models (Admin) — Connect AI providers such as Anthropic, OpenAI, Google, Azure, or AWS
  2. Connect Data (Admin / Builder) — Register databases, data lakes, files, or timelines
  3. Create Custom Models (Admin / Builder) — Define AI assistants that combine a provider model with data sources
  4. Trigger Semantic Extraction (Admin / Builder) — Build the semantic layer from the connected data
  5. Start a Session (All Users) — Open a session, select a custom model, and query your data

Guide Areas​

AI Agent​

Interact with the Lakehousecat AI Agent using natural language. The Agent generates charts, manages platform resources, and executes analytical tasks — with available capabilities determined by your role (User, Builder, or Admin).

Available to: All users (scope varies by role)

Sessions​

Have analytical conversations with the AI. Ask questions in natural language — the AI queries your data through the semantic layer and returns answers, charts, and insights. Covers the input toolbar, file uploads, voice input, prompt shortcuts, chart selection, session history, and sharing.

Available to: All users

Data​

Connect and manage external data connections. Covers all supported types (PostgreSQL, MySQL, MSSQL, Delta Lake, Apache Hudi, Apache Iceberg, S3, File Upload, Timeline), the creation workflow, advanced settings, and the data source editor.

Available to: Admin, Builder

Models​

Create and manage AI model configurations. Covers Provider Models (credentials and API connections for each supported provider) and Custom Models (purpose-built AI assistants linked to data sources).

Available to: Provider Models — Admin only | Custom Models — Admin, Builder

Analytics​

Work with charts and dashboards generated from AI-assisted data analysis. Charts are produced during sessions and can be reviewed, approved, shared, and embedded in dashboards.

Available to: All users

Operations​

Automate data workflows using job definitions. Covers creating job definitions with tasks, scheduling via cron syntax, manual trigger, and monitoring job run history.

Available to: Admin, Builder

Prompts​

Save frequently used queries as reusable prompts. Prompts are invoked with a /command shortcut in the session input and can be shared with other users.

Available to: All users