SQL in Laminar

Laminar uses SQL as the primary language for defining stream transformations. If you know SQL, you already know how to build streaming pipelines.

All examples in this documentation use the sample data.

The Pattern

Every Laminar pipeline follows a simple pattern:

INSERT INTO destination
SELECT transform(columns)
FROM source
WHERE conditions

That's it. You read from a source, transform the data with SQL, and write to a destination.

A Complete Example

Let's calculate revenue per region from the orders stream.

INSERT INTO region_metrics
SELECT
  region,
  TUMBLE(interval '5 minutes') as window,
  COUNT(*) as order_count,
  SUM(amount) as revenue
FROM orders
WHERE status = 'completed'
GROUP BY region, window

region	window_start	window_end	order_count	revenue
us-east	2024-01-15 10:00:00	2024-01-15 10:05:00	3	364.99
us-west	2024-01-15 10:00:00	2024-01-15 10:05:00	2	375.50

The query filtered to completed orders, grouped by region and 5-minute windows, and computed aggregates.

What Makes Streaming SQL Different?

In traditional SQL, tables are finite. In streaming SQL, data never stops arriving.

-- This would never complete in streaming!
SELECT COUNT(*) FROM events
 
-- Instead, use time windows:
SELECT
  TUMBLE(interval '1 minute') as window,
  COUNT(*) as event_count
FROM events
GROUP BY window

Results emit every minute as windows close:

window_start	window_end	event_count
10:00:00	10:01:00	8
10:01:00	10:02:00	12
10:02:00	10:03:00	5

Key Concepts

Time Windows

Since streams are unbounded, you define time boundaries for aggregations.

SELECT
  TUMBLE(interval '1 minute') as window,
  SUM(amount) as total
FROM orders
GROUP BY window

window_start	window_end	total
10:00:00	10:01:00	225.50
10:01:00	10:02:00	250.00
10:02:00	10:03:00	425.00

SQL Engine

Laminar's SQL is powered by Apache DataFusion, giving you:

Standard SQL syntax
100+ built-in functions
High performance Rust execution
Streaming-specific extensions for windows and time

Next Steps

Quick Start - Build your first SQL pipeline
Sample Data - Reference dataset for all examples
Data Types - Supported types and how to use them
Time & Windows - Master streaming time concepts
Functions - All available SQL functions