dbt for on‑chain data

dbt is a tool data teams use to turn raw tables into cleaned, tested datasets with version control. Dune said it now lets those dbt workflows run on blockchain data and send the results straight into Snowflake or BigQuery. (docs.dune.com) Dune described the release as “dbt connector → Datashare” in a blog post published last week. The company said the flow starts with Dune’s onchain catalog, runs dbt models on DuneSQL, and then syncs the output into a configured warehouse target. (dune.com) In Dune’s documentation, the feature is limited to configured Datashare targets such as Snowflake or BigQuery. The docs say teams can keep transformation logic inside Dune while consuming the finished tables inside their own warehouse. (docs.dune.com) The product is aimed at a common data-engineering problem: blockchain data is large, messy, and expensive to reshape after export. Dune said its connector lets teams apply business logic, filter protocols, and aggregate data before anything is synced out. (dune.com) That changes where the work happens. Instead of exporting raw chain data first and building a second pipeline in a warehouse, teams can model data on Dune and deliver only the tables they want to maintain downstream. (docs.dune.com) Dune’s docs say the dbt connector uses a write-enabled Trino endpoint, which is a SQL access layer that lets dbt create and update tables in a private namespace. The same docs distinguish it from Dune’s read-only Trino connector for business-intelligence tools and ad hoc querying. (docs.dune.com) The company also says the setup fits existing engineering workflows. Its product page lists compatibility with dbt projects, Apache Airflow workflows, Prefect flows, incremental queries, and Git-based review and deployment. (dune.com) Dune is pitching the connector as part of a broader enterprise stack around Datashare. Its Datashare page says customers can access Dune data in Snowflake, BigQuery, and Databricks, while the new dbt-to-Datashare docs currently describe publishing transformed outputs to Snowflake or BigQuery. (dune.com, docs.dune.com) The release is now listed in Dune’s changelog with incremental sync support, post-hook automation, and SQL-based lifecycle management. That leaves Dune closer to the way many analytics teams already work: build models in dbt, test them in code, and ship curated tables into the warehouse they already use. (docs.dune.com)