Agents can now instantly consume and reuse artifacts generated by CLAIRE GPT. This enables you to ask questions about the artifacts instantaneously. Artifacts can include files such as PowerPoint presentations, reports, or PDFs, as well as Intelligent Data Management Cloud assets such as mappings or rule specifications. Using this enhancement, you can access, manage, and reuse artefacts, and promote greater efficiency and collaboration.
This release includes the following data quality agent specific enhancements:
Generate data quality rule for a data set
CLAIRE GPT introduces the capability for the data quality agent to generate data quality rule specifications from a file that contains sample data, or a data set from a specified catalog source. CLAIRE GPT analyzes the critical data elements of a data set to generate rule specifications.
Simplified process steps
When you ask CLAIRE GPT to generate a data quality rule specification or cleanse data, the process steps are presented concisely. The steps now help you distinguish the required steps from optional ones and from steps where you can choose to provide additional inputs. Additionally, if you choose to skip an optional step, the response explicitly explains the impact of skipping that step.
Improved explanation for rule generation
When you ask CLAIRE GPT to generate a data quality rule specification, the Reasoning section in the response now clearly explains the rationale behind how the data quality rule was generated. This includes adding comments to the rule scripts that CLAIRE GPT generates as a part of the rule generation plan.
Improved data quality rule recommendations
CLAIRE GPT now considers business context from the user prompt to identify relevant data elements and recommend data quality rules that are contextual to the identified data elements. Additionally, if no rule specifications are available in Data Quality for CLAIRE GPT to recommend for the data set, CLAIRE GPT can generate new rules for the data set.
Improved data quality assessment
When you use CLAIRE GPT to profile data, you can now prompt to analyze a subset of a data set. For example, you can ask CLAIRE GPT to focus on a sample row within the data set, analyze consistent and repeatable subsets of a data set, or to examine a representative subset.
Additionally, you can now ask CLAIRE GPT to profile up to 10,000 rows and 50 columns of data.
For more information about the data quality agent, see Data Quality Agent.
Enhanced product help agent with discovery context awareness
The product help agent now retains the context from the discovery agent to provide relevant help responses. When you ask help questions related to a discovered asset, the product help agent considers the previous context of the asset while answering the question.
For example, if you ask questions to discover specific assets and then ask a help question related to the discovered asset, the product help agent retrieves asset details and identifies the asset type. It then customizes its response based on the identified asset type and returns the most relevant response.
This enhancement improves user experience by delivering context-aware, asset-specific help, reducing the need for repetitive questions in a conversation.