⚡️
Declarative state
Describe pipelines, jobs and tables as desired outcomes instead of step-by-step scripts.
Databricks, declarative.
Declarative control for Databricks
Describe the Databricks workspace you want, not the steps to get there. Mortar handles the synchronization so you can ship without the grind.
Features
Everything you need to manage Databricks workspaces like Infrastructure as Code, with clear plan and apply steps.
🔒
Guardrails built in
Enforce access, tagging, and guardrails automatically so teams move fast without risk.
📊
Plan then apply
See every change before it lands with readable plans and clean, predictable outcomes.
🧩
Reusable config
Package workspace patterns as modules so teams can launch new projects in minutes.
How Mortar works
Configuration-driven control for enterprise Databricks environments.
You declare
version: 1
pipeline:
name: "orders_daily_ingest"
mode: "batch"
source:
type: "file"
path: "abfss://raw/sales/orders/{env}/"
target:
bronze_table: "sales_orders_raw"
silver_table: "sales_orders_clean"
partition_by:
- "order_date"
schedule:
type: "daily"
time: "02:00"
compute:
job_cluster:
spark_version: "14.3.x-scala2.12"
autoscale:
min_workers: 2
max_workers: 8
schema_policy:
mode: "evolve_additive"
primary_key:
- "order_id"
Mortar delivers
Mortar
Plan
Apply
Databricks
Job clusters
Autoscaling job cluster (2-8 workers) pinned to your Spark version.
Tables
Bronze sales_orders_raw and silver sales_orders_clean created with partitions on order_date.
Pipelines
Batch pipeline orders_daily_ingest wired with source, target, and partitioning from the spec.
Schedules enforced
Daily 02:00 UTC job with governance-friendly schema policy (evolve_additive).
Get in Touch
Want clearer governance and lower Databricks costs? Start the conversation.