> ## Documentation Index
> Fetch the complete documentation index at: https://parabola.io/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# Excluding processed records
In this lesson, you'll learn how to use a Parabola Table as reference data so your flow only processes new records — and never re-processes data it's already handled in a prior run.
***
Add a branch that detects when any category's Total Sales has changed since the last run — so your team is automatically notified when the numbers shift.
Press **S** to open the search bar, search for **Pull from Parabola Table**, and drag it onto the canvas as a standalone step — don't connect it to anything yet. Place it above the **Combine data & calculate total sales by category** card, to the left of your **Send to Parabola Table** step. Configure it to pull from your **Sales by Category** table.
```text wrap theme={null}
I want to add a new branch off of my "Sum total sales by category" step.
First, compare it against my "Pull from Parabola Table" step to find any categories where "Total Sales" has changed.
Then, send an email notification with those changed category rows included.
```
Since the CSV data is static, the values will match the previous run and the branch will show zero changes — which is the correct behavior. If the underlying sales data were updated, any category with a new total would be caught and sent automatically.
Drag the email step up so it lives above the **Generate CSV file** step on the canvas, keeping the flow visually organized with the notification branch clearly separated from the export path.
Select all three steps, then right-click and add them to a card. Optionally add an emoji and color to the card.
The steps added to your canvas and their exact documentation may differ from what you see below — every AI response is unique. The important thing is that your resulting data set matches what you see here.
