DOCUMENTATION
Everything you need to get started with DVT. Most operations have a point-and-click way — the hub, the connection manager, Martin, the scheduler — and these docs teach that way first, with the terminal way right behind it for people who live there.
QUICKSTART
Install DVT and run your first cross-engine model in 5 minutes.
INSTALL & PLATFORMS
The supported-platform matrix — macOS, Linux x86_64, Python 3.10–3.14 — and the WSL2 path for Windows.
DBT COMPATIBILITY
Why '100% compatible' holds: DVT runs your installed dbt-core — feature-by-feature proof.
BENCHMARKS
The methodology behind every number we publish — real runs, real engines, one laptop.
CLI COMMANDS
Full reference for all DVT commands: run, build, seed, test, sync, debug, docs, login.
PROFILES.YML
Database connections, target environments, environment switches vs engine shifts.
SOURCES.YML
Bind every source to its engine with meta.connection — the key to federated reads and painless target switches.
F-TABLE & F-INCREMENTAL
The crown jewel: federated materializations that join any engines and land results anywhere.
EXECUTION PATHS
How DVT decides between pushdown, extraction, and cross-target materialization.
SUPPORTED DATABASES
15 verified-or-supported adapters, bucket storages, file formats, and Python-model sources.
LEAST-PRIVILEGE PREREQUISITES
What to grant DVT on every engine, per use case — and the list of things it never asks for. The page to hand your DBA.
BUCKET STORAGES
Use S3, GCS, Azure Blob, and local filesystem as extraction sources or materialization targets.
ADAPTER GUIDES
Per-adapter setup guides with profiles.yml examples, configuration fields, and driver details.