Building and Using Google Cloud AI and ML Solutions CLOUD-DIGITAL-LEADER Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   An insurance company records all customer service calls for quality assurance. They want to automatically identify calls where the customer expressed frustration and extract the main reason for the call (e.g., 'claim status', 'billing issue'). The team wants a fully managed solution with no custom ML training. What should they use?

   • ATrain a custom AutoML model on the audio files to detect frustration
   • BUse the Speech-to-Text API to transcribe calls, then the Natural Language API for sentiment analysis and entity extraction
   • CUse the Text-to-Speech API to generate audio summaries of the calls
   • DUse the Vision API to scan call transcripts and classify them
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A logistics company stores shipment-delay records in BigQuery. Their data analysts are proficient in SQL but have no Python or ML training. They want to build a model that predicts whether a shipment will be delayed based on origin, destination, weather, and historical performance. The model must be created and run entirely within their existing data warehouse environment. What should they do?

   • AWrite a Python script using TensorFlow and run it on a Cloud TPU
   • BUse the Natural Language API to analyze the shipment records and predict delays
   • CExport the data to CSV and train a custom model using AutoML Tables
   • DCreate and train a model using BigQuery ML with a CREATE MODEL statement in standard SQL
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A bank's mobile app wants to add a feature that reads account balances and recent transactions aloud for visually impaired customers. The text is already available in the app. The feature must work offline-free, be fully managed, and require no ML expertise. What should they use?

   • AUse the Natural Language API to summarize the transactions into a shorter text
   • BUse the Speech-to-Text API to capture voice commands from the user
   • CUse the Vision API to render the text as an image for the screen reader
   • DUse the Text-to-Speech API to convert the balance and transaction text into natural-sounding audio
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   A global e-commerce company wants to automatically translate product descriptions into multiple languages for its international customers. Which Google Cloud API is most appropriate?

   • ANatural Language API
   • BSpeech-to-Text API
   • CCloud Translation API
   • DVision API
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A healthcare software company wants to build a custom AI assistant that answers patient questions using their proprietary clinical guidelines and historical interaction data. They need fine-grained control over the model's behavior, the ability to ground responses in their own documents, and the flexibility to integrate with their existing patient portal. They have a team of ML engineers. Which approach best meets these requirements?

   • AUse the pre-trained Natural Language API for entity extraction and sentiment analysis on patient queries
   • BUse the Cloud Translation API to translate patient queries and return translated responses
   • CUse AutoML to train a custom classification model on the historical interaction data
   • DBuild a custom solution using the Gemini Enterprise Agent Platform (Vertex AI) with their own data
   Reveal Solution Discussion Select an answer first