GitHub

Uploading Custom Vector Data

GeoRetina AI allows you to upload your own geospatial data and query it using natural language, just like you would with public data sources. This feature enables you to bring your proprietary data into GRAI and interact with it seamlessly.

Supported Data Formats

You can upload vector data in several common formats:

FormatFile ExtensionNotes
GeoJSON.geojson, .json

Preferred format for web applications

Shapefile

.zip (containing .shp, .dbf, .shx, etc.)

Must be uploaded as a ZIP archive with all component files

KML/KMZ.kml, .kmz

Common format from Google Earth applications

Uploading Your Data

1. Access the Data Upload Options

When chatting with GRAI, you have several ways to upload geospatial data:

Geospatial data attachment options

2. Select and Upload Your File

  1. Click the upload button or drag and drop your file
  2. Select the geospatial data file from your computer
  3. Wait for the upload to complete

Once processing is complete, you'll see a confirmation that your data has been successfully imported:

Data successfully imported as queryable layer

During the upload process, you'll see a progress indicator:

Upload process indicator

3. Data Processing and Import

After uploading, GRAI will process your data to make it queryable:

  1. The system validates the file format and structure
  2. The data is imported and indexed for fast querying
  3. Attributes and feature types are analyzed
  4. The data becomes available as a queryable layer

Viewing Uploaded Data

You can see all your uploaded data layers in the data layers panel:

Uploaded data shown in layers list

Querying Your Uploaded Data

Once your data is uploaded, you can query it using natural language, just as you would with any other data source:

Querying an uploaded data layer

Example queries for your uploaded data:

"Show me all features in my uploaded layer"
"Count the number of points in my data that are within the city boundaries"
"What is the total area of polygons in my uploaded dataset?"
"Show me features in my uploaded layer where the 'status' field is 'completed'"

Best Practices for Custom Data

For optimal results when working with your own data:

  • Clean your data before uploading to ensure quality results
  • Use descriptive attribute names to make querying more intuitive
  • Keep data size reasonable - very large datasets may take longer to process
  • Include relevant attributes that you might want to query
  • Use appropriate coordinate systems - GRAI works best with WGS84 (EPSG:4326)
  • Verify attribute types are correct (numeric fields should be numbers, not text)

Data Privacy and Security

Your uploaded data remains private and secure:

  • Files are encrypted during transit and storage
  • Data is only accessible to your account or organization
  • You can delete uploaded data at any time
  • No data is shared with third parties

Limitations

Current limitations to be aware of:

  • Maximum file size: 50MB per upload
  • Maximum number of features: 100,000 per dataset
  • Temporal data support is limited
  • Complex attribute relationships may not be fully queryable

Combining with Public Data

One of the most powerful aspects of GRAI is the ability to combine your uploaded data with public datasets:

"Show me OpenStreetMap schools that are within 500m of points in my uploaded dataset"
"Compare the density of features in my dataset with the population density from Overture data"
"Find areas where my uploaded polygons overlap with protected natural areas"

This capability allows you to gain insights that wouldn't be possible with either dataset alone.