Clean and prepare data by making field values consistent and
standardized across your dataset.
This agent can perform the following tasks:
- Normalize fields (for example, country names, codes, city
names)
- Standardize date formats
- Standardize and verify address fields
- Apply custom coding for advanced normalization (for example,
metro/non-metro classification)
- Ensure data correctness against business rules
-
Select or create a pipeline (for example,
Property_Policy_Holders).
-
In the pipeline canvas, click Gio™ AI Assistant to view its
panel.
-
Select one of the following options to start the
agent:
- Normalize Fields: For making
field values consistent (for example, fixing typos,
standardizing codes).
- Standardize Fields: For
enforcing a single, correct format (for example,
date formats).
- Enrich Spatial Data: For
adding spatial information (optional).
-
Type your request in the input box so the agent can
analyze the dataset and suggest recommended steps.
For example, type: "I want to standardize the Date,
Country, state, addresses fields in this dataset."
-
Review the Recommended Steps
flowchart and a detailed Steps and
Reasoning table displayed by the
agent.
Each step includes a reasoning column that explains its
purpose, as shown in the following examples:
- Standardize Field: Converts
country values to 3-letter ISO codes, flags unmapped
values.
- Standardize Date: Parses and
outputs dates in a single format
(
YYYY-MM-DD), flags invalid or
out-of-range values.
- Verify Address: Standardizes
and validates address fields for consistency and
accuracy.
- Custom Coding: Normalizes
city names, classifies metro/non-metro, flags
unmatched entries.
-
Click Add Steps to insert the
recommended steps into your pipeline.
-
Optional: You can ask the agent to add further
steps, such as: "Help me to add a step to ensure correctness
of the data against business rules."
The agent creates a new suggestion that includes the
additional steps and the earlier recommendation (for example,
additional standardization or validation).
Your quality pipeline now includes normalization and
standardization steps configured by the agent. You can run the
pipeline to clean and standardize your dataset.
Use cases
- Country normalization: Converts
"INDIA", "india", "IND" to "IND" (ISO3), flags unmapped
values.
- Date standardization: Converts
"15/03/1985", "22-07-90", "1988/11/08" to
YYYY-MM-DD.
- Address verification: Standardizes
and validates address fields for matching and
enrichment.
- Business rule validation: Adds steps
to check data correctness against defined business
rules.