AI-103 Practice Question 69: Language Detection Ticket Routing
Microsoft Azure AI Apps and Agents Developer Associate (AI-103) · Test 7 · question 9 of 10
A support platform receives tickets in whatever language the customer typed, with no language field set by the submission form. Before a ticket reaches Custom Translator or a language-specific analysis model, the platform needs to identify which language each ticket is written in, including telling apart Bosnian, Croatian, and Serbian text written in Latin script versus Serbian written in Cyrillic, so it is routed to the right downstream pipeline.
Which Azure AI Language feature should sit at the front of the pipeline?
Show answer and explanation
Answer:
Language detection is the purpose-built feature for this: it "can identify more than 100 languages in their primary script," and separately offers script detection, returning "a script name and script code according to the ISO 15924 standard" for languages written in more than one script; Microsoft names Kazakh as an example, and the same multi-script situation applies to Serbian's Latin/Cyrillic usage. That combination of language plus script is exactly what the ticket router needs to pick a downstream pipeline correctly.
- B is wrong because PII detection redacts sensitive spans in text; identifying the document's language isn't its purpose, and bolting language identification onto it isn't how the service is designed to be used. - C is wrong because key phrase extraction assumes the text's language is already known (or auto-detected internally) to extract phrases meaningfully; it isn't a language identifier, and inferring language from extracted-phrase vocabulary is not a supported or reliable technique. - D is wrong because it puts a translation-time setting in the position of an upstream classification step. The ticket needs to be *routed* based on its language before it reaches Custom Translator or an analysis model; Custom Translator translating a source language it auto-detects internally doesn't help the platform choose which pipeline to send the ticket to in the first place.