GriidAi LogoGriidAiTMHelp Center
Ask GriidAi

Workspace and screen-aware AI assistant

How GriidAi reads your active screen environment—including map extent, zoom level, visible layers, selected features, and open tables—so you can use conversational shortcuts without manual parameter entry.

The GriidAi assistant is aware of your real-time workspace state and what is active on your screen. Because the assistant sees what you see, you no longer need to type out long, explicit parameters, coordinates, or redundant layer specifications.

Real-time screen and workspace awareness

When you submit a request in the chat composer, the assistant inspects your active screen environment to understand the context of your question.

The assistant reads the following workspace elements in real time:

  • Map view and zoom level: The geographic bounding box currently displayed on your map canvas.
  • Visible layers: The specific vector and raster layers currently toggled to visible in your Layers catalog.
  • Selected features: Vector features or records actively selected in the map or attribute table.
  • Open attribute tables: Any dataset tables currently opened in the Table Drawer.

Workspace context available to the assistant

Workspace contextWhat it enables
Map view & zoomSpatial operations such as clipping, buffering, or bounding based on the area currently visible on screen
Visible layersCommands referring to visible data without needing to re-type exact layer names
Selected featuresTargeted summaries, filtering, and statistical analysis on active feature selections
Open tablesContextual field editing, column renaming, filtering, and record inspection

Conversational shortcuts

Because the assistant understands your active screen context, you can use natural conversational shortcuts instead of typing manual bounding coordinates or complex queries:

Map and viewport operations

Clip this to the current view.

Extracts active spatial layers to the precise bounding box currently displayed on your screen canvas.

Zoom to that layer.

Directs the map viewport straight to the geographic extent of the specified dataset.

Show me what is visible in the current map view.

Inventories or summarizes only the features and layers currently in view.

Selection-based analysis

Summarize the parcels I've selected.

Generates statistical breakdowns and count metrics for the actively highlighted vector features.

Calculate statistics for the selected features.

Computes mean, median, sum, or distributions exclusively for the selected subset of records.

Export the selected features to a new layer.

Isolates the active selection and saves it as an independent catalog layer.

Layer and table context

Work with the visible layer.

Applies operations to whichever layer is currently toggled visible on the canvas.

Summarize the data in the open table.

Generates descriptive statistics for the dataset currently open in the Table Drawer.

Rename 'col_1' to 'Building_Height'.

Updates field metadata directly in the open attribute table without recreating the dataset.

Best practices for screen-aware prompts

  1. Check visible layers before querying: If multiple layers overlap in your map view, toggle off layers you do not want included in view-based operations.
  2. Frame your viewport: Zoom and pan the map to frame the exact area of interest before issuing commands like "Clip this to the current view".
  3. Verify feature selections: Ensure only the intended features are highlighted before requesting selection-based summaries.
  4. Review the proposed plan: GriidAi always presents a structured plan showing detected inputs, bounding areas, and methodology before executing modifications. Always verify these parameters before confirming.

Need assistance with this workflow?

Our geospatial engineering team is available to help with custom datasets, STAC queries, and algorithms.

Get Support

On this page