Describe features of Natural Language Processing workloads on Azure

The correct answer is Azure AI Language with sentiment analysis.

This is the best choice because the telecommunications company needs to:
1. Evaluate customer satisfaction - which requires understanding the emotional tone of text
2. Categorize support tickets by emotional tone - which is exactly what sentiment analysis does

Sentiment analysis is a feature of Azure AI Language service that analyzes text and returns sentiment labels (such as positive, negative, neutral, or mixed) along with confidence scores. This directly addresses the requirement to evaluate customer satisfaction from support ticket descriptions and categorize them by emotional tone.

Why the other options are not suitable:

Named entity recognition and key phrase extraction combined would identify specific entities (like names, locations, organizations) and important phrases in the text, but these features do not evaluate emotional tone or customer satisfaction. They extract factual information rather than analyzing sentiment.

Azure AI Speech with pronunciation assessment is designed for evaluating spoken language pronunciation quality, typically used in language learning scenarios. Since the company is working with support ticket descriptions (text), not voice recordings, and needs emotional tone analysis rather than pronunciation evaluation, this service is inappropriate.

Conversational language understanding is designed to extract intent and entities from conversational queries to build chatbots and command interfaces. While useful for understanding what action a customer wants to take, it doesn't specifically analyze or categorize emotional tone or satisfaction levels, which is the core requirement here.

The correct answer is Azure Speech Services text-to-speech API.

This service is specifically designed to convert written text into natural-sounding spoken audio. It uses neural text-to-speech technology to generate human-like speech from text input, making it ideal for applications like e-learning platforms, accessibility tools, virtual assistants, and content narration. The text-to-speech API supports multiple languages, voices, and customization options including speech synthesis markup language (SSML) for fine-tuning pronunciation, pace, and tone.

Why the other options are incorrect:

Speaker recognition API with voice profile configuration is designed for identifying or verifying speakers based on their unique voice characteristics. It analyzes audio to determine who is speaking, not to generate speech from text. This is used for authentication and identification purposes.

Language identification API with audio synthesis extension is not a real service offering. Language identification is used to detect which language is being spoken in audio input, and there is no standard 'audio synthesis extension' that converts text to speech.

Speech-to-text API with inverse transcription mode is incorrect because speech-to-text converts spoken audio into written text (the opposite of what's needed). There is no 'inverse transcription mode' in Azure Speech Services - for converting text to speech, you would use the dedicated text-to-speech API instead.

The correct answer is Speech synthesis.

Speech synthesis is the technical term for the Azure Speech Services capability that converts written text into natural-sounding audio output. This process is also commonly known as text-to-speech (TTS). Speech synthesis takes written text as input and produces spoken audio as output, using advanced AI models to generate human-like voices with natural intonation, pronunciation, and rhythm.

Why the other options are incorrect:

Speech recognition is the opposite process - it converts spoken audio into written text, not text into audio. This is also known as speech-to-text (STT).

Speech translation with audio rendering is a more complex feature that involves translating spoken language from one language to another and then rendering it as audio. While this does include audio output, it's not the specific technical term for the basic text-to-audio conversion capability.

Text transcription refers to the process of converting audio recordings into written text, which is essentially the same as speech recognition. This is the reverse of what the question asks about - it produces text output rather than audio output.