Data Quality Agent
Purpose
Data Quality Agent supports Solvency II data preparation by building a partial data dictionary from client-provided Excel metadata. It maps Data IDs to attribute names, descriptions, datasets or worksheets, formats, and source file names using the information provided in client files and supporting reference sources.
Primary users
The primary user is specified as “Both.” In practice, the agent is intended for the relevant users involved in the Solvency II data dictionary workstream, although the specific business or technical roles are not further specified in the provided information.
Where it fits (process/stage/trigger)
The agent fits into a Solvency II data documentation process when a client provides Excel metadata containing Data IDs, example values, file names, and worksheet names. It is triggered by the availability of client-provided files that need to be transformed into a structured partial data dictionary.
Key capabilities / workflow
The agent analyzes client Excel metadata, checks whether Data IDs are available, maps those IDs to attribute names and related metadata, validates whether the dictionary coverage is sufficient, and enriches the mapping using the internal “Data Dictionary Example” collection and web research when needed. It then generates a structured output table for review or downstream use.
Inputs
Typical inputs include client-provided Excel metadata, Data IDs, examples of values, file names, worksheet names, client-provided files, the internal “Data Dictionary Example” collection, and web research sources.
Outputs / Deliverables
The output is a table containing Data IDs, attribute names, descriptions, datasets, formats, and storage file names. This table represents the partial data dictionary built from the client-provided Excel metadata and supporting reference sources.
Value
Data Quality Agent helps accelerate the creation of a structured Solvency II data dictionary from fragmented Excel metadata. It improves consistency in mapping Data IDs to business-readable attributes and supports clearer documentation of dataset, format, and source file information.
