Data Quality Agent > Data quality rule generation > Sample conversation
  

Sample conversation

Let's take a look at a sample conversation to help you get started on your Data Governance and Catalog data quality rule generation journey in CLAIRE GPT.
As a data analyst in a financial organization, you are responsible for ensuring the accuracy of the country codes for the regions in which your organization operates. You have a file that contains rule descriptions, and you ask the agent to generate rule specifications from the rule descriptions.
To achieve the goal, perform the following steps:
  1. 1Log in to CLAIRE GPT.
  2. 2Upload a file that contains the rule descriptions, and start a conversation to identify how the data quality agent can generate data quality rules from the rule descriptions saved in a file.
    3. Enter the following prompt:
    Generate rules from this file.
    CLAIRE GPT indicates the optional steps within the plan. By default, CLAIRE GPT executes all steps, but you can prompt it to skip the optional ones.
    The following image shows the response displaying a detailed plan on how the agent can generate rules from a file:
    The response displays steps on how the data quality agent uses the data quality rule generation skill to generate rule specifications from your file.
  3. 3If you are happy with the proposed rule generation plan, enter the following prompt:
    4. Yes, proceed with the plan.
    The data quality rule generation skill of the agent validates the rules and checks for existing rules with the same logic in Data Quality. The agent groups rules with logically identical descriptions into clusters and creates a single rule for each cluster in the file. If CLAIRE GPT can't find the specified dictionary in Data Quality, it automatically creates the dictionary for the rule specification.
    The agent then generates rule scripts and corresponding test data for each rule, executes tests for each rule, and provides a comprehensive summary of the generated rule specifications.
    The following image shows the data quality rules that the agent generated from the uploaded file.The response displays rules that the agent generated based on business requirements and data patterns from the file you uploaded.
    To see the details of each generated rule specification, you can switch to Detail View.
    You can download the data as a CSV file or copy the data to another file.
  4. 4To view a summary of the rule specifications that the agent generated, click the Data Quality Rules Agent Log card.
    5. The following image shows the canvas with a summary of the rule specifications that the agent generated.The response displays a summary of generated rules, distribution of rules across different data quality dimensions, and rules that the agent could not generate.
  5. 5If you are happy with the rule specifications that the agent generated and want to publish them in Data Quality, enter the following prompt:
    6. Publish Rules
    To view the published rule specifications, click the Published DQ Rules card. If a published rule fails a test, the Published DQ Rules card indicates that.
    You can open the rule specifications in Data Quality and view them in the CLAIRE folder within the CLAIRE project.
Using the data quality rule generation skill of the data quality agent, CLAIRE GPT has analyzed the rule description in an uploaded file, extracted rules from the file, generated rule specifications, and published them in Data Quality.