ADA-C01 Question 430
Single answerMonitor AI usage and costsA Snowflake administrator is asked to review rising spend related to Snowflake AI features and identify which teams are driving the cost increase. The company wants a solution that provides account-level visibility into AI service consumption over time and supports chargeback analysis by user or workload. Which approach should the administrator use?
- A
Query ACCOUNT_USAGE views that expose AI service usage and costs, then correlate that data with query history, users, roles, or tags to attribute consumption to teams
- B
Review only warehouse credit consumption in WAREHOUSE_METERING_HISTORY, because all Snowflake AI feature usage is fully included in warehouse credits
- C
Use STORAGE_USAGE_HISTORY to identify which schemas are generating the highest AI costs, because AI billing is recorded as database storage growth
- D
Rely exclusively on RESOURCE_MONITORS, because they directly track and break down Snowflake AI service consumption by feature and by user
Show answer and explanation
Correct answer: A
Explanation
The best answer is to use Snowflake's account-level usage and billing metadata for AI-related services and combine it with activity metadata for attribution. In practice, administrators monitor consumption through Snowflake billing/usage surfaces such as Account Usage and Snowsight, then correlate with QUERY_HISTORY and organizational constructs such as users, roles, warehouses, applications, or tags to support internal reporting and chargeback. This is the most defensible administrative approach because AI features may incur charges outside standard warehouse metering. By contrast, WAREHOUSE_METERING_HISTORY and RESOURCE_MONITORS focus on warehouse compute, and STORAGE_USAGE_HISTORY is unrelated to AI service billing. Snowflake documentation on Account Usage, organization/account cost monitoring, and AI feature billing best supports this pattern.
- A. Correct.
Correct. For monitoring AI usage and cost, the appropriate administrative approach is to use Snowflake-provided usage and billing telemetry in Account Usage/Snowsight and then correlate that information with operational metadata such as query history, users, roles, warehouses, or object tags for internal attribution and chargeback. This aligns with real-world administration patterns: first identify the billed AI service usage at the account level, then map activity back to consuming teams or workloads.
- B. Incorrect.
Incorrect. This is a common misconception. Not all Snowflake AI-related consumption is represented purely as warehouse credit usage. Some AI features are billed separately from virtual warehouse compute, so looking only at WAREHOUSE_METERING_HISTORY would miss or understate actual AI service costs.
- C. Incorrect.
Incorrect. STORAGE_USAGE_HISTORY is intended for storage consumption trends and related charges, not for attributing AI service usage. AI service spend is not determined by schema storage growth, so this would not provide an accurate basis for monitoring or chargeback.
- D. Incorrect.
Incorrect. Resource monitors are useful for controlling or alerting on virtual warehouse credit usage, but they do not provide a direct, feature-level breakdown of Snowflake AI service consumption by user. Using them alone would not satisfy the requirement to monitor AI usage and attribute rising AI costs across teams.