Notebooks

Notebooks are your private workspace for data exploration. They work like Jupyter notebooks—code cells, markdown cells, and outputs—all running in the cloud.

Creating Notebooks

From the Notebooks page:

1. Click New notebook
2. Choose one of:
   • Blank notebook – Start fresh with an empty code cell
   • Upload .ipynb – Import an existing Jupyter notebook

Your notebooks are private by default and auto-save as you work.

The Kernel

The kernel is your Python execution environment. It auto-connects when you open a notebook, so you can start running code immediately.

How It Works

• The kernel connects automatically in the background
• If you run a cell before the kernel is ready, it queues and executes once connected
• The status button in the toolbar shows the current state (connecting, warming up, ready)

Kernel Timeouts

The kernel has built-in limits to keep your session secure and responsive:

Kernel Actions

• Restart kernel – Clear all variables and start fresh (via the ⋮ menu)
• Disconnect – End the session manually (click the green kernel button)
• Clear queue – Cancel any pending cell executions (via the ⋮ menu)

Working with Cells

Cell Types

Adding Cells

• Click Add cell at the bottom of the notebook
• Press B to add below the current cell
• Press A to add above the current cell

Running Code

• Shift + Enter – Run current cell and advance to the next
• Ctrl/Cmd + Enter – Run all code cells in order
• Click the Run all button in the toolbar

Cells can be run even while another cell is executing—they queue up and run in order, just like Jupyter. If the kernel isn't ready yet, cells wait and execute once it connects.

Cell Navigation

When a cell is focused (blue border):

Cell Actions

Outputs

When you run a code cell, outputs appear below:

• Text – Print statements and returned values
• Tables – DataFrames render as formatted tables
• Charts – Matplotlib, Plotly, and other visualizations
• Images – Any PIL/Image output
• Errors – Tracebacks with syntax highlighting

Collapsing Outputs

Long outputs can be collapsed:

• Click the collapse button on the output
• Press O when the cell is focused

Clearing Outputs

• Click Clear outputs in the toolbar to clear all cell outputs
• Individual outputs clear when you re-run the cell

Code Completion

When the kernel is connected, you get intelligent code completion:

• Completions appear as you type
• Press Tab or Enter to accept
• Press Escape to dismiss

Inspector

Get documentation for any symbol:

1. Place your cursor on a function or variable
2. Press Ctrl/Cmd + I
3. A tooltip shows the docstring

Saving

Notebooks auto-save continuously. The status indicator shows:

Press Ctrl/Cmd + S to force an immediate save.

Importing & Exporting

Importing Notebooks

You can import existing Jupyter notebooks with full output preservation:

1. Click New notebook
2. Choose Upload .ipynb
3. Preview the imported content
4. Click Save to confirm

Margin imports:

• All code and markdown cells
• Cell outputs including charts, images, tables, and errors
• Execution counts and the notebook title

Exporting Notebooks

Download your work as a standard Jupyter notebook:

1. Open the notebook you want to export
2. Click the ⋮ menu in the toolbar
3. Select Download as .ipynb

The exported file works in JupyterLab, Google Colab, Kaggle, and any other Jupyter-compatible environment.

Next Steps

• Work with the AI agent to accelerate your analysis
• Create briefs from your notebook outputs
• Upload datasets to analyze
• Share your work with public links