> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wherobots.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Give Waldo your first prompt

> Open Waldo in Wherobots Cloud, give it a prompt in plain language, and get a table, a map, and the code behind the answer.

**Waldo** is the AI agent in Wherobots Cloud for spatial data engineering, Earth observation, and geospatial analytics. You give Waldo a prompt in plain language. Waldo finds the datasets in your catalogs, writes and runs the code on WherobotsDB and RasterFlow, and returns a summary, tables, and a map. Nothing to install: Waldo runs in your browser.

> [!NOTE]
>
> **Waldo is in public preview.** It's available in every Organization Edition, including the free Professional Edition trial. Features and limits may change before general availability.

## Before you begin

* A Wherobots Account in any Organization Edition. New Organizations start on a free Professional Edition trial. To create one, see [Create a Wherobots Account](https://docs.wherobots.com/get-started/wherobots-cloud/create-account.md).
* To analyze your own data, add it to a catalog first. See [Data Catalogs](https://docs.wherobots.com/get-started/initial-storage/data-hub.md). Without your own data, Waldo uses Wherobots Open Data, such as Overture Maps and US Census boundaries.

> [!TIP]
>
> If you don't see **Waldo** in the left navigation, an Admin in your Organization has turned it off. Ask them to turn it on in **Organization Settings** > **Preferences**, or start with [Start Your First Notebook](https://docs.wherobots.com/get-started/wherobots-cloud/start-notebook.md) instead.

## Give Waldo a prompt

**1. Open Waldo**

Log in to [Wherobots Cloud](https://cloud.wherobots.com/). In the left navigation, click **Waldo**.

![The Ask Waldo page with suggested prompts and the Ask about your data box](https://mintcdn.com/wherobots/pIL7XBACyDQjslj7/images/develop/waldo/empty-state.png?fit=max\&auto=format\&n=pIL7XBACyDQjslj7\&q=85\&s=e50f016a1fc42e09373557f39a836b30)

The Waldo start page

**2. Write your prompt**

Type a prompt in the **Ask about your data** box and press **Enter**. Name the place, what you want measured, and the output you want. Try one of these:

* In King County, Washington, where are trees posing availability risks to power infrastructure? This joins a canopy height raster to power line geometries in WherobotsDB.
* Which H3 level 5 cells contain the most building footprints per square kilometer? This is a global vector aggregation in WherobotsDB.
* Which homes across Mesa, Arizona have rooftop solar and which don't? RasterFlow detects panels in aerial imagery, and WherobotsDB joins them to building footprints.

**3. Answer clarifying questions**

If the prompt can be read more than one way, Waldo asks up to three questions and suggests a default for each. Answer them, or reply "use your defaults".

**4. Approve compute**

Waldo reads catalogs and data without asking. Before it starts a runtime, runs a billed task such as a RasterFlow task, or changes a file, an approval card appears in the chat. Click **Approve** to continue.

**5. Read the answer**

The answer starts with the result in one or two sentences, then tables, a map card, and the datasets and method Waldo used. Click **Open the map** to view the results in Wherobots Map. Click **Worked** to see each step and the code behind it.

![A Waldo answer for Mesa, Arizona rooftop solar, with a map card, a summary table, and an inline map](https://mintcdn.com/wherobots/cznO35op89Akau7K/images/develop/waldo/answer-map-card.png?fit=max\&auto=format\&n=cznO35op89Akau7K\&q=85\&s=f54d84a7abb15ecea49a49bfbae75d90)

A finished answer with a map card, a summary table, and an inline map

**6. Refine it**

Keep asking in the same chat, for example "Show only the rural tracts" or "Save the results as GeoParquet in My Files".

## Next steps

> [card] [Work with Waldo](https://docs.wherobots.com/develop/waldo/use-waldo.md)
>
> Approvals, follow-ups, files, tips for better questions, and how to check an answer.

> [card] [Waldo overview](https://docs.wherobots.com/develop/waldo/overview.md)
>
> What Waldo is best at, what it can't do yet, and what it costs.

> [card] [Schedule recurring tasks](https://docs.wherobots.com/develop/waldo/schedules.md)
>
> Turn a Waldo script into a task schedule.

> [card] [Start your first notebook](https://docs.wherobots.com/get-started/wherobots-cloud/start-notebook.md)
>
> Write and run spatial SQL and Python yourself in a Notebook Runtime.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.