Python Analysis Cells for Live Dashboard Analytics

Python Analysis Cells

A Python Analysis Cell is a chart card that runs a Python script against your live datasource data and renders the result as an interactive ApexCharts visual. Think Jupyter notebook — but embedded directly in your dashboard, connected to your real data, and decorated with AI-generated insights automatically.

Plan requirement: Python Analysis Cells are available on Enterprise plans.

How it works

1. Add a new chart and set the type to Python.
2. AI Insights fetches a DataFrame from your datasource using the standard query configuration.
3. Your script receives the data as a df variable (a pandas DataFrame).
4. Your script returns a result dict describing the chart to render.
5. AI Insights renders the result as an ApexCharts visual and runs the AI narrative automatically.

All Python execution happens in the browser via Pyodide (Python compiled to WebAssembly). No server-side Python runtime is required.

Writing your first script

# df is pre-loaded from your datasource — no connection code needed.
import pandas as pd

monthly = df.groupby('Month')['Revenue'].sum().reset_index()
monthly['MA3'] = monthly['Revenue'].rolling(3).mean()

result = {
    "chartType": "line",
    "title": "Monthly Revenue with 3-Month Moving Average",
    "labels": monthly['Month'].tolist(),
    "values": monthly['Revenue'].tolist(),
    "series": [
        {"name": "Revenue",    "data": monthly['Revenue'].tolist()},
        {"name": "3-Month MA", "data": monthly['MA3'].fillna(0).tolist()}
    ]
}

Click ▶ Run to execute. The chart renders immediately.

The result object

Available libraries

Limitations