Hex Magic AI

Prompt Magic to join 'Subscriptions' with 'User_Events'.

Ask Magic to identify cohorts with declining activity.

Use Magic Charts to visualize the trend.

Publish as a secure app.

Select a SQL result dataframe.

Prompt Magic: 'Forecast the next 12 months using Prophet based on this dataframe.'

Magic generates the Python code and installs necessary libraries.

The analyst reviews and tweaks the confidence intervals via NL.

Connect both sources to Hex.

Use Magic to write a cross-database comparison script.

Identify row count mismatches automatically.

Generate a report highlighting the specific discrepant IDs.

Highlight all cells in the notebook.

Trigger Magic 'Explain Notebook'.

The AI generates markdown headers and cell-level comments.

The AI suggests improved variable names for readability.

Pull ticket text via SQL.

Use Magic to write a Python script using a HuggingFace model.

Categorize tickets into 'Positive', 'Neutral', and 'Negative'.

Aggregate results by product feature automatically.

Ask Magic to create a first-touch and last-touch attribution model.

Prompt for a comparison between the two models.

Use Magic to generate a table of 'Under-valued channels'.

Export the resulting logic to a dbt model.

Connect to a live data stream.

Use Magic to define 'normal' bounds using standard deviation logic.

Set up an automated alert cell that triggers via Magic when bounds are exceeded.

Integrate with Slack for instant notifications.