Snowflake Metadata

This page describes what Snowflake metadata Espresso reads and why. In general, we look at workload metadata to be able to understand and optimize your traffic patterns.

Espresso will never access your Snowflake data.

We read metadata from Snowflake's ACCOUNT_USAGE and, when available, ORGANIZATION_USAGE schemas. The regular product and the savings estimate use different sets of metadata. Some views depend on your Snowflake edition or enabled features; when a view is unavailable, we skip it.

Most customers, including public companies, are comfortable sharing this data under NDA.

Savings estimate

For a savings estimate, we pull only the metadata below.

Snowflake view	Why we use it
QUERY_HISTORY	To understand what workloads you're running and when, along with statistics such as runtime, queue time, and bytes scanned.
QUERY_ATTRIBUTION_HISTORY and QUERY_METERING_HISTORY	To understand compute cost at the query level.
WAREHOUSE_EVENTS_HISTORY	To see warehouse events, such as warehouses turning on and off and clusters spinning up and down.
WAREHOUSE_METERING_HISTORY	To measure hourly credit usage by warehouse.
METERING_DAILY_HISTORY	To measure daily credit usage across the account.
RATE_SHEET_DAILY, CONTRACT_ITEMS, and REMAINING_BALANCE_DAILY	To calculate financial projections and understand committed spend.

We also run SHOW WAREHOUSES to read warehouse settings such as size and type. We use Snowflake's CURRENT_REGION, CURRENT_ACCOUNT, CURRENT_ORGANIZATION_NAME, and CURRENT_ACCOUNT_NAME functions to uniquely identify the account.

QUERY_HISTORY contains user-generated query text. If your query text is sensitive, we're happy to sign an NDA or a BAA before sharing data. We can remove query text from a savings estimate export upon request. All other fields in these views are machine-generated.

We can also omit RATE_SHEET_DAILY and CONTRACT_ITEMS from a savings estimate upon request.

Regular product

The regular product pulls the savings estimate metadata above, plus the views below. This broader set lets us report on all Snowflake spend, understand account configuration, and identify cost outliers and edge cases. Many of these views are feature-specific and optional; unavailable views are skipped.

Category	Additional Snowflake views
Workload and warehouse activity	ACCESS_HISTORY, SESSIONS, WAREHOUSE_CLUSTER_UTILIZATION, and WAREHOUSE_LOAD_HISTORY (for accounts where warehouse-load collection is enabled)
Billing, services, and data movement	APPLICATION_DAILY_USAGE_HISTORY, AUTOMATIC_CLUSTERING_HISTORY, CATALOG_LINKED_DATABASE_USAGE_HISTORY, COPY_FILES_HISTORY, DATA_QUALITY_MONITORING_USAGE_HISTORY, DATA_TRANSFER_HISTORY, DATABASE_REPLICATION_USAGE_HISTORY, HYBRID_TABLE_USAGE_HISTORY, ICEBERG_STORAGE_OPTIMIZATION_HISTORY, INTERNAL_DATA_TRANSFER_HISTORY, LISTING_AUTO_FULFILLMENT_USAGE_HISTORY, MARKETPLACE_PAID_USAGE_DAILY, MATERIALIZED_VIEW_REFRESH_HISTORY, METERING_HISTORY, NOTEBOOKS_CONTAINER_RUNTIME_HISTORY, OPENFLOW_USAGE_HISTORY, PIPE_USAGE_HISTORY, QUERY_ACCELERATION_HISTORY, REPLICATION_GROUP_USAGE_HISTORY, SEARCH_OPTIMIZATION_HISTORY, SERVERLESS_ALERT_HISTORY, SERVERLESS_TASK_HISTORY, SNOWFLAKE_APP_RUNTIME_COMPUTE_HISTORY, SNOWPARK_CONTAINER_SERVICES_HISTORY, SNOWPIPE_STREAMING_CLIENT_HISTORY, SNOWPIPE_STREAMING_FILE_MIGRATION_HISTORY, and USAGE_IN_CURRENCY_DAILY
Storage and backup	ARCHIVE_STORAGE_DATA_RETRIEVAL_USAGE_HISTORY, BACKUP_OPERATION_HISTORY, BACKUP_SETS, BACKUP_STORAGE_USAGE, BLOCK_STORAGE_HISTORY, DATABASE_STORAGE_USAGE_HISTORY, POSTGRES_STORAGE_USAGE_HISTORY, STAGE_STORAGE_USAGE_HISTORY, STORAGE_LIFECYCLE_POLICY_HISTORY, STORAGE_REQUEST_HISTORY, STORAGE_USAGE, and TABLE_STORAGE_METRICS
Catalog and security snapshots	COMPUTE_POOLS, DATABASES, SCHEMATA, TABLES, VIEWS, TASKS, PIPES, REPLICATION_GROUPS, SERVICES, USERS, GRANTS_TO_USERS, GRANTS_TO_ROLES, and NETWORK_POLICIES
Cortex and AI usage	CORTEX_AGENT_USAGE_HISTORY, CORTEX_AI_FUNCTIONS_USAGE_HISTORY, CORTEX_AI_GUARDRAILS_USAGE_HISTORY, CORTEX_AISQL_USAGE_HISTORY, CORTEX_ANALYST_USAGE_HISTORY, CORTEX_CODE_CLI_USAGE_HISTORY, CORTEX_CODE_DESKTOP_USAGE_HISTORY, CORTEX_CODE_SNOWSIGHT_USAGE_HISTORY, CORTEX_DOCUMENT_PROCESSING_USAGE_HISTORY, CORTEX_FINE_TUNING_USAGE_HISTORY, CORTEX_FUNCTIONS_QUERY_USAGE_HISTORY, CORTEX_FUNCTIONS_USAGE_HISTORY, CORTEX_PROVISIONED_THROUGHPUT_USAGE_HISTORY, CORTEX_REST_API_USAGE_HISTORY, CORTEX_SEARCH_BATCH_QUERY_USAGE_HISTORY, CORTEX_SEARCH_DAILY_USAGE_HISTORY, CORTEX_SEARCH_SERVING_USAGE_HISTORY, DOCUMENT_AI_USAGE_HISTORY, and SNOWFLAKE_INTELLIGENCE_USAGE_HISTORY

We take daily snapshots of catalog and security views so we can understand how the account changes over time. We also read warehouse settings through Snowflake's warehouse API.