Boilstream Quickstart

Build a real-time streaming pipeline in under 15 minutes.

Prerequisites

Docker (20.10+) and Docker Compose
A message source (Kafka, or use the built-in test generator)

Installation

# Binary
curl -fsSL https://get.fabricent.com/boilstream | sh

# Docker
docker pull fabricent/boilstream:latest

# Homebrew
brew install fabricentlabs/tap/boilstream

Create a Pipeline

Create pipeline.yaml:

name: orders-pipeline

source:
  type: kafka
  config:
    brokers: ["localhost:9092"]
    topic: raw-orders
    group: boilstream-orders

transforms:
  - name: filter-valid
    sql: |
      SELECT * FROM input
      WHERE amount > 0 AND status IS NOT NULL

  - name: enrich
    sql: |
      SELECT
        *,
        CASE
          WHEN amount > 1000 THEN 'high-value'
          WHEN amount > 100 THEN 'medium-value'
          ELSE 'standard'
        END AS tier
      FROM input

sink:
  type: postgresql
  config:
    connection: postgresql://user:pass@localhost:5432/analytics
    table: processed_orders

Run the Pipeline

# Start processing
boilstream run pipeline.yaml

# With built-in test data generator
boilstream run pipeline.yaml --generate-test-data

# Check status
boilstream status

Monitor

Boilstream exposes Prometheus metrics at :9090/metrics and a health check at :9090/health.

Next Steps

Add multiple sources and sinks
Configure exactly-once delivery guarantees
Set up dead-letter queues for failed events
Deploy with Docker Compose or Kubernetes

Need 24/7 streaming ops? Our Managed Service keeps your pipelines running reliably around the clock.