Transforming with AI
Learn about Parabola’s AI transformation steps designed to categorize, standardize, and extract information from large datasets.
Building challenge
To the right of the card titled “Calculate revenue by row and assign statuses”, add a new card titled “Categorize products with AI” and add a Categorize with AI step to the card.
Assign each value in the Product Title column into one of the following categories:
- Pants
- Tops
- Hats
- Other
Check your work ▼
Leveraging AI Transformation Steps in Parabola
Beyond automating manual processes, many customers use Parabola to improve workflows and leverage AI. We've already covered a few ways AI is integrated into Parabola, such as:
✅ Parsing PDFs to extract structured data
✅ Automatically configuring step settings with AI
In addition to these, Parabola offers a suite of AI-powered transformation steps that can enhance your workflow.
AI-Powered Steps in Parabola
1. Categorize with AI
This step assigns rows to custom categories based on text analysis.
- Example: Categorizing product titles into Pants, Tops, Hats, etc.
- AI processes each row and assigns it to the appropriate category dynamically.
2. Extract with AI
Converts unstructured text into structured data.
- Example: Extracting city, state, and zip code from a messy text field.
3. Standardize with AI
Ensures consistency across variations of the same term.
- Example: Standardizing "DHL Expedited," "DHL Express," and "DHL" into a single, uniform value.
4. Experiment with AI
Allows you to custom-prompt AI within your workflow—similar to using ChatGPT directly inside Parabola.
Exploring AI Steps in Parabola
To see all available AI-powered steps, simply search "AI" in the step search bar. New AI transformations are continuously being added.
Example: Categorizing Products with AI
In this example, we have a Product Title column and want to categorize products into groups:
- Add the Categorize with AI step.
- Define categories: Pants, Tops, Hats, Other.
- AI processes the data and assigns each row to a category.
- The results appear in a new column with clean, structured labels.
If you need to fine-tune the AI's categorization, use the "Fine-Tuning" section to add custom prompts for better accuracy.
Try It Yourself!
Test the Categorize with AI step in the building challenge below, and let us know if you have any questions! 🚀