Skip to main content
Use the Wherobots Model Context Protocol (MCP) server to go from no prior schema knowledge of your spatial data to production-ready spatial analysis in minutes. No schema memorization or manual SQL debugging.

​
Automated catalog discovery (0:30)

The Problem: You just joined a team with terabytes of spatial data across dozens of catalogs. Where do you even start? The Solution: Just ask. Let our MCP server navigate your data landscape. The user poses a simple question:
“What taxi data do I have in Wherobots?”
The agent identifies NYC Yellow Taxi data with 185 million rows, complete with schema definitions and column types. This eliminates hours of manual catalog browsing and documentation hunting.

​
Natural Language Spatial Queries (3:13)

The Problem: Writing spatial SQL requires knowing exact table names, coordinate systems, spatial functions, and data formats. Make one mistake and you’ll spend the next hour debugging instead of analyzing your results. The Solution: The agent demonstrates Contextual Inference. Watch what happens when the user simply asks:
“Find me all the burger joints near the space needle in Seattle”.
Notice what’s NOT in that prompt:
  • ❌ Coordinates
  • ❌ Polygons
  • ❌ Search radius
  • ❌ Dataset names
  • ❌ SQL syntax
Here are the key capabilities at play:
  • Intelligence: The agent infers what is needed and selects the Overture Maps dataset and constructs a spatial join.
  • Self-Correcting: When the initial query fails (wrong column format), the agent debugs itself—re-reading the schema and fixing the SQL without any human intervention.
  • Precision: Final results show burger restaurant locations with exact distances from the Space Needle.

​
Generating Production Notebooks (4:45)

The Problem: Chat is great for quick answers, but production work needs notebooks, scripts, and repeatable workflows. Copy-pasting from chat to code is error-prone and time-consuming. The Solution: Instruct the agent to:
“Create a Jupyter notebook using examples from Wherobots’ docs which finds all the burger joints that are within one mile from a fire station. Check that all the queries will work at a small scale before using in the notebook.”
Why This Matters: The agent runs validation tests inside the chat, catching errors before they make it into the notebook. No broken code artifacts. The Result: A production-ready .ipynb file with tested, working spatial SQL—ready to run at scale in Wherobots Cloud.
Use the MCP server to accelerate your geospatial workflows from experimentation to production.

​
Execution in Wherobots Cloud (9:08)

The Problem: Production workflows need repeatable, shareable notebooks rather than one-off chat queries. Translating chat explorations into production artifacts usually requires manual rewriting and testing. The Solution: The notebook generated by the MCP agent can be uploaded directly to Wherobots Cloud and runs without modification. All queries—whether executed through chat or in notebooks—run on Wherobots Cloud infrastructure. Watch how the agent’s validated logic transitions from chat to notebook:
  • From Chat to Notebook: What started as an interactive conversation becomes a documented, repeatable workflow.
  • Same Infrastructure: Both the queries generated in chat and the notebook execute on a Wherobots Cloud runtime.
  • Production Ready: The notebook generated by the MCP server processes spatial joins across the entire US—analyzing distances between thousands of fire stations and burger joints using the exact logic the agent defined and validated earlier.

​
Ready to try it yourself?

The Wherobots MCP server transforms spatial data exploration from a manual, error-prone process into an intelligent, automated workflow. Whether you’re discovering datasets, writing complex spatial queries, or generating production notebooks, the agent handles the complexity so you can focus on insights. Install the MCP server today and use it to accelerate your geospatial workflows. All powered by the Model Context Protocol and Wherobots Cloud.