#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-engineer-interview-prep/dbt-and-the-transformation-layer/incremental-models

░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░█████████████████████████████████████████████████████████████████████████████████████████████████████████████

#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-engineer-interview-prep/dbt-and-the-transformation-layer/incremental-models

Repovive

Incremental Models - dbt and the Transformation Layer | Data Engineer Interview Prep | Repovive

Repovive.

Data Engineer Interview Prep16 sections · 140 units8h total

Open in Course

Incremental Models

Processing only new data

Incremental models only process new data. Critical for large tables.

Basic pattern:

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

SELECT * FROM {{ source('raw', 'events') }}
{% if is_incremental() %}
  WHERE event_time > (
    SELECT MAX(event_time) FROM {{ this }}
  )
{% endif %}

Strategies:

append: Just add new rows
merge: Update existing + add new (requires unique_key)

Caution: Incremental models can drift from source. Run full refreshes periodically.