dbt | data build tool

Data Modeling for Teams

dbt makes analyst teams more efficient by leveraging software development principles.



dbt is a free and open-source data modeling tool designed for growth-stage companies who want to spend less time fussing with tooling, and more time exploring their data.


How it Works


1. Setup a project for your team

Projects use git flow to enable collaboration. Once you install dbt, run dbt init, and push that to a Github repository. Now, your whole team can access your data models.


2. Develop models in isolation

Each analyst can use dbt to build their own isolated copy of the analytics schema. There's no need for your analysts to step on each others' toes any longer.


3. Populate your warehouse

Use Sinter or cron to keep your production data warehouse running the latest and greatest version of your models.





Leverage the power of SQL

dbt models are built as SQL selects -- allowing you to express your transformations in the same way you explore your data. Once you have written these selects, dbt takes care of materializing them in your warehouse.


events_by_user.sql

with users as (
  select * from {{ref('users')}}
), events as (
  select * from {{ref('events')}}
)

select
 users.id as user_id,
 count(*) as num_events

from users

left join events on events.user_id = users.id

group by users.id





Learn More


Documentation

Take a look at the dbt documentation and getting started guide.

Mailing List

Join the mailing list to get periodic updates on dbt.