Episode Details

Agents dominated dbt Summit, but does smart orchestration or a context layer make data agent-ready on its own? The episode argues no. Fanatics Betting and Gaming saw about 0.2% model reuse from dbt State until it declared source freshness and model-level SLAs, which raised reuse to roughly 15%. The same pattern runs through Fivetran's Context Layer, AI audit trails and token costs.

Listeners learn how dbt State is billed per daily active target table, including tests, why the Context Layer's beta status needs checking in writing, how git history serves as an audit trail, and how dbt models served through the MCP server cut Gong transcript tokens by 20x or more.

Key takeaways

• Declare source freshness and model-level SLAs before expecting dbt State to reuse models and cut warehouse costs.
• Get the release status and pricing basis of Fivetran's Context Layer in writing before you evaluate it.
• Write metric definitions with semantic name, owner, source models, grain, filters, time basis, unit rules, freshness and exclusions.
• Keep transformation logic in git so you can show a diff when the CFO asks why a number was revised.
• Pick the metric your AI pilot is asked about most and give it a named owner, written definition, freshness tolerance and one test before Friday.

Chapters
0:00 Fanatics and the 0.2% dbt State reuse rate
2:21 dbt State pricing and per-test charges
3:53 Fivetran Context Layer: early access risks
6:26 Auditing AI agents without a rulebook
7:49 Why token costs belong to data teams
9:17 Declare meaning before agents act

Go deeper, free lessons

• dbt (data build tool): industrialized SQL transformation
• The metrics & semantic layer
• Active metadata and the data catalog
• Data FinOps: controlling cloud data cost
• Shift-left data quality: embedding governance in the engineering pipeline

Full article and transcript: https://www.mba-training.com/blog/agent-ready-data-dbt-summit

MBA Training, mba-training.com