Datenentitäten

In addition to the documentation of the application MODUS M365 Data Entity Processor, some special features for data preparation in connection with MAGiC are briefly explained here.

Data entities allow the creation of JSON structure data from any combination of tables. However, we have found that a simple structure without many child entities is easier for an AI to analyze. Ideally, therefore, keep the defined structures as flat and, with regard to costs, as small as possible.

You can also show a Demo Document in the definition of the Data Entities (JSON). It makes sense to have this data analyzed on a test basis in the chat area of the AI (Playground → Chat): Analyze the following data and describe what you can see..

This will give you a good impression of whether the AI can interpret and use the data well.

Example

Analyze the following data and describe what you can see.

{"DocumentType":"Return Order","DocumentNumber":"1001","CustomerNumber":"10000","Name":"Möbel-Meller KG","Address":"Tischlerstr. 4-10","City":"Düsseldorf","Contact":"Herr Michael Emanuel","PostCode":"48436","OrderDate":"2025-01-23","LanguageCode":"DEU","Salesperson":"Detlef Hembrock","Email":"herr.michael.emanuel@contoso.com","$^$Parent$^$":null}

Answer:

Based on the given data, it appears to be a return order for a company called "Möbel-Meller KG" with a document number of 1001. The customer number is 10000 and the address is "Tischlerstr. 4-10" in the city of Düsseldorf, with a postcode of 48436. The contact person for this order is "Herr Michael Emanuel" and the order date is January 23, 2025. The language code is "DEU" which stands for German. The salesperson for this order is "Detlef Hembrock" and the email associated with the order is herr.michael.emanuel@contoso.com. Finally, the "\(Parent\)" field is null, which indicates that this order is not related to any other orders or documents.