Question 130
A production Snowflake Gen AI workflow must handle the following: An insurer classifies claim narratives into document-needed queues. Which implementation choice most directly resolves the stated requirement?
Explanation
Correct Answer
“Use task-specific text classification against the approved category set and test both boundary cases and unsupported labels” is supported by this requirement: An insurer classifies claim narratives into document-needed queues. Fixed-category classification requires a defined label set and performance assessment by class. Verify it using per-class precision/recall with confusion analysis for the approved taxonomy, including ambiguous inputs.
Incorrect Answers
“Use general generation without any fixed category contract or label checks” would not meet the requirement that an insurer classifies claim narratives into document-needed queues. Fixed-category classification requires a defined label set and performance assessment by class. The more reliable approach is to use task-specific text classification against the approved category set and test both boundary cases and unsupported labels.
“Use TRANSLATE and map output length to category names” would not meet the requirement that an insurer classifies claim narratives into document-needed queues. Fixed-category classification requires a defined label set and performance assessment by class. The more reliable approach is to use task-specific text classification against the approved category set and test both boundary cases and unsupported labels.
“Use SENTIMENT as the only classifier for topic categories that are not emotional” would not meet the requirement that an insurer classifies claim narratives into document-needed queues. Fixed-category classification requires a defined label set and performance assessment by class. The more reliable approach is to use task-specific text classification against the approved category set and test both boundary cases and unsupported labels.