Any Source. Any Target. One SQL Project.

DVT is a cross-engine data transformation tool built on dbt. Write SQL models referencing any database — DVT handles extraction, federation, and loading automatically.

GET STARTEDpip install dvt-core
models/cross_engine_report.sql
-- This model joins MySQL CRM data with Snowflake ERP data
-- DVT handles extraction automatically

{{ config(materialized='table') }}

SELECT
    c.customer_name,
    c.email,
    o.order_date,
    o.total_amount
FROM {{ source('crm', 'customers') }} c        -- MySQL
JOIN {{ source('erp', 'orders') }} o            -- Snowflake
    ON c.customer_id = o.customer_id
WHERE o.order_date >= '2024-01-01'

WHY DVT

🏠

OWN YOUR DATA

Run on YOUR hardware. No cloud dependency. No vendor lock-in. Your data stays where YOU put it.

💰

CUT COSTS

Eliminate ETL tool licenses ($50K-500K/yr). Reduce cloud compute spend. One tool replaces three.

🚀

FASTEST INSIGHTS

From raw source to business dashboard in minutes. SQL you already know. No learning curve.

🔒

ON-PREM FIRST

Purpose-built for your infrastructure. Cloud is a source, not a requirement. Scale on your terms.

HOW IT WORKS

1

WRITE SQL

Reference any source on any database using standard dbt syntax. ref(), source(), config() — all the same.

2

DVT ROUTES

Smart path selection: pushdown to the target engine when possible, extract via Sling + DuckDB when cross-engine.

3

DATA LANDS

Results materialize in your target database. Tables, views, incremental — all supported across engines.

FEATURES

🔗

CROSS-ENGINE SQL

Write SQL models that JOIN data from any database. MySQL + Snowflake + PostgreSQL in one SELECT.

🔄

DBT COMPATIBLE

100% backward compatible. Same ref(), source(), config(). Existing dbt projects work unchanged.

💻

ZERO INFRASTRUCTURE

Runs entirely on your laptop. No clusters, no cloud services, no infra to manage.

SMART PUSHDOWN

When all sources are on one engine, SQL pushes down directly. Zero data movement, full adapter speed.

📈

INCREMENTAL EXTRACTION

Delta extraction with watermarks. Only new/changed rows move. Persistent DuckDB cache between runs.

📦

19 ADAPTERS, ONE PACKAGE

PostgreSQL, Snowflake, BigQuery, Databricks, MySQL, Oracle, SQL Server — all built in.

19 ENGINES. ONE PACKAGE.

All adapters included in dvt-adapters. No separate installs.

PostgreSQLSnowflakeBigQueryRedshiftDatabricksDuckDBMySQLMariaDBSQL ServerOracleSparkFabric

READY TO START?

Install DVT in seconds. Your existing dbt project works out of the box.