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

# Creating a Notebook

> Create an analysis notebook with the assistant, a blank notebook, a template, or an uploaded file

Analysis notebooks combine Python code, explanations, and results. They run in Atomscale with access to your organization's data through the preinstalled Python SDK.

## Prerequisites

Sign in to the correct organization. You need access to the data you want to analyze. Assistant features are not available to demo accounts; these accounts can start with a blank notebook, available templates, or an uploaded notebook.

## Create with the assistant

**1. Open Analysis**

Select **Analysis** in the section switcher, then **New Notebook** in the left sidebar.

**2. Describe the investigation**

Enter your request in **Describe what you want to analyze**. Include the growth or data item name and your engineering question.

**3. Create the notebook**

Click **Create Notebook**. Atomscale opens the notebook and sends your request to its assistant when the editing session connects. The assistant can add analysis code and explanations to the notebook.

**4. Inspect and refine**

Review the notebook and continue the conversation in the right sidebar's **Agent** panel. See [Working in a Notebook](https://docs.atomscale.ai/platform/reference/notebooks/working.md).

For an example investigation, see [Analyze](https://docs.atomscale.ai/platform/get-started/analyze.md).

## Start from a blank notebook or template

On **New Notebook**, select **Blank Notebook** to start from the SDK setup. Select an available organization template to create a separate notebook from it. You can also browse **Templates** in the Analysis sidebar. See [Reusable Templates](https://docs.atomscale.ai/platform/reference/notebooks/templates.md).

The available templates depend on what your organization has created. A guide's example does not require a template with the same name.

## Upload an existing notebook

**1. Open the upload dialog**

On **New Notebook**, click **Upload Existing Notebook**.

**2. Choose a file**

Select one file. Supported formats are a marimo notebook (`.py`), a Jupyter notebook (`.ipynb`), a plain Python script (`.py`), or a Markdown notebook (`.md`).

**3. Review and upload**

Review the selected file and notebook name, then click **Upload**. Formats other than a marimo notebook are converted to marimo before the notebook opens.

Jupyter outputs and attachments are not preserved during import. Run the imported analysis to produce results in Atomscale. Conversion may require changes to code, dependencies, or references to local files. See [Troubleshooting](https://docs.atomscale.ai/platform/reference/notebooks/troubleshooting.md#import-fails).

## Next

* [Working in a Notebook](https://docs.atomscale.ai/platform/reference/notebooks/working.md)
* [Saving, Versions, and Access](https://docs.atomscale.ai/platform/reference/notebooks/saving-and-access.md)


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