Models

Models serve as structured representations of datasets within Workplace AI. Models enable users of the system to understand and query structured data sources - such as databases, spreadsheets, and other tabular formats - alongside unstructured data such as documents, emails and multimedia.

Administrators can define the schema, or model attributes, which specifies the structure, fields, and relationships within the dataset. This schema acts as a blueprint, ensuring that data loaded from various sources—such as CSV, XML, JSON, or ODBC connectors - is consistently interpreted and integrated into the platform.

Models can be used to standardise data integration by ensuring data from disparate sources is mapped to a common structure and because the data model can be aligned to your use case, structured data represented by models can be queried directly, allowing users to perform advanced searches, organise and visualise the data.

There are two types of model in Workplace AI:

• Models - Are the top level data structures that can be used. Models represent tables, SQL tables or top level objects e.g. an Invoice or Purchase Order

• Sub Models - For complex data structures sub models can be used to represent nested tables or sub objects. Sub Models cannot have further sub models. A Sub models can be re-used within multiple models e.g. an Organisation or Line item which can be used within a Purchase order or Invoice model.

Throughout this documentation, the use case of a Customer, Invoice (Model) and Line Item (Sub Model) will be used as a worked example.