Getting Started with GeoRetina AI

This guide walks through the standard GeoRetina AI workflow: add a site, boundary, or asset context; choose the right agent mode; ask a focused risk or conservation question; inspect the resulting layers; and reuse generated assets for follow-up work.

1. Start from the Chat Composer

The composer is where you provide instructions and choose the context for the run. The most important controls are:

• Add: Attach files, draw a site or region of interest, import from the database, or open custom actions.
• Core/Super: Choose the agent mode.
• Assets: Select regions, asset inventories, exposure layers, analysis layers, and reports from the current chat or project.
• Project selector: Keep work organized inside the right project.
• Chat settings and Integrations: Configure the chat and connect external providers.

2. Add Site or Asset Context

Open Add when you need to attach a file, draw an ROI, or bring in an existing asset.

Common starting points:

• Attach files for local asset inventories, boundaries, field observations, or documents needed in the current chat.
• Select ROI on map to draw a polygon region of interest.
• Database to import saved regions, asset layers, exposure layers, analysis layers, or other project assets.
• Actions to manage custom slash actions when available for your account.

3. Choose Core or Super Mode

Use Core for fast map-layer and metric outputs. Use Super for more advanced, in-depth analysis. Super can create deeper report assets, dynamic visualizations, and follow-up-ready context. It can also automatically combine raster outputs with vector assets when that helps explain risk, exposure, or conservation patterns.

4. Ask a Focused Question

Good first prompts include:

• The analysis type.
• The region or layer to analyze.
• The time period or comparison window.
• Any output preference, such as map, table, report, or exported data.

Use @ when you want to select a saved region or layer exactly instead of retyping its name.

5. Inspect the Map Layers

Results appear as map layers and artifacts. Use the layer controls to adjust opacity, inspect legends, open tables, export data, or remove layers from the map.

If the result includes a large exposure or asset output, open the attribute table to review features, sort fields, and inspect records.

6. Use the Artifacts Panel and Insights Viewer

The artifacts panel summarizes project context, generated risk or conservation layers, and reports. Use it to return to previous outputs or confirm which assets are available for follow-up prompts.

For richer report outputs, open the Insights viewer from the chat response.

7. Reuse Results

After a run completes, use the Assets control to reuse regions, asset inventories, exposure layers, analysis layers, and reports in follow-up messages.

Follow-up examples:

• "Of the OSM features in the physics-based flood exposure layer, summarize the items in the highest risk class."
• "Use this vegetation stress layer and create a short field-inspection brief."
• "Compare the 2025 built-area growth to the previous land-cover trend output."
• "Explain the dominant hazard driver in this physical-risk screening."

8. Switch Between 2D and 3D When Useful

The map is the canonical 2D review surface for risk and conservation outputs. When a 3D view is useful for communicating terrain, buildings, or urban context, use the Map/3D toggle. See 3D Map View for the full workflow.

Next Steps

• Learn prompt structure in Effective Prompting.
• Organize work in Workspace Organization.
• Run your first Land Use/Land Cover Mapping workflow.
• Set up External Data Integrations.