GenAI Model Input Documentation Generator

Purpose

The purpose of GenAI Model Input Documentation Generator is to generate structured documentation from metadata and model code, including data dictionaries, variable descriptions, and lineage notes.

Primary users

The primary user is not specified. The agent is associated with AQS and is owned by Ronan Davit.

Where it fits (process/stage/trigger)

This agent fits into model documentation activities where model metadata, transformation logs, variable dictionaries, and model code are available and need to be converted into structured documentation.

Key capabilities / workflow

The agent extracts information from the provided metadata and model code, checks whether the available sources are sufficient, generates structured documentation, validates whether the documentation structure is complete, and refines gaps before delivering the documentation.

Inputs

User-provided inputs are not specified. Available datasets include model metadata, transformation logs, and variable dictionaries, and the use case also references metadata and model code.

Outputs / Deliverables

The specified deliverables are structured documentation, including a data dictionary, variable descriptions, and lineage notes. Other output formats are not specified.

Value

The agent helps create structured model input documentation from existing model-related sources, supporting clearer understanding of variables, transformations, and lineage in an insurance context.