From data exploration to migration
Run Otto in your terminal or directly in Astro. Otto has access to your instance, your warehouses and your deployment history.
ORCHESTRATION FOR DATA AND AI.
Move data and agentic workflows from prototype to production.
Powered by Apache Airflow®, augmented with AI.
BUILT
WITH
WITH
Featured Content
Booking.com delivers travelers a connected trip with data and AI
Read Case Study →
Janus Henderson Investors built self-healing pipelines with Astro & Otto.
Read Case Study →
Astro: Airflow engineered for speed and scale.
We rebuilt Airflow's scheduler, executor, runtime, and infrastructure for greater scale, lower latency, and faster recovery. Your DAGs stay the same, while Astro runs low-latency pipelines, AI workflows, and hundreds of thousands of tasks at once.
- 500,000sustained concurrent tasks25× open-source Airflow
- 230msp95 task start latency103× faster than Celery at 2× the load
- <15 min RPOcross-region recoveryRTO under 1 hour
Coordination
Airflow API Server1-10 replicas
Airflow Triggererup to 8 replicas
Airflow Dag Processor2 replicas with HA
Executionone executor per deployment
Airflow Celery Executorautoscaled workers
Airflow Kubernetes Executorone pod per task
Platform
Astro Scheduler
01 / 05
Event-driven scheduling that stays fast
The Astro Scheduler reacts to new work as it arrives instead of waiting for a polling cycle. Task starts stay fast as trigger volume and concurrency grow.
Run event-driven pipelines and AI workflows alongside ETL, ML, and batch workloads—all on one Airflow platform.
p95 task start latency: Celery has 23.582 seconds of task lag at 50,000 concurrent tasks; Astro at 100,000 takes 230 milliseconds; and Astro at 300,000 takes 300 milliseconds.
Introducing Otto
THE ONLY AGENT BUILT FOR APACHE AIRFLOW.
Astronomer’s Airflow expertise, plus your team’s conventions. Otto writes Dags, investigates failures, and plans upgrades grounded in both.
Learn MoreOtto reasons with your context
Otto plans, executes, and verifies its work against public Airflow knowledge, Astronomer's proprietary expertise, and your private memory.
Every action sharpens the loop
Every fix, pattern, and correction feeds back into your private memory. The more your team uses Otto, the smarter it gets.
Raise the ceiling of data engineering with Airflow you never have to babysit
You write DAGs, Astro runs them reliably at any scale. No Kubernetes expertise required. No ops burden. Just Airflow that works the way it should.
$ brew install astro
✓ Successfully installed astro-cli v1.39.0
$ astro dev init --from-template etl
Initializing Astro project...
- ✓ Created project directory
- ✓ Generated example DAG
- ✓ Project ready!
Run `astro dev start` to run Airflow locally
$ astro deploy
- Building image...
- Deploying to Astro...
- ✓ Deployment complete!
View at: cloud.astronomer.io
