Amazon MWAA brings managed Airflow to AWS-native teams who want to stop running Airflow themselves. Kestra takes a different approach: declarative YAML orchestration that runs in any language on any cloud, covering data pipelines, infrastructure automation, and business processes, without requiring a VPC, S3 bucket, or IAM role before your first workflow runs.
Declarative YAML orchestration that runs on Docker, Kubernetes, or any cloud, in any language: Python, SQL, Bash, R, Go, Node.js. No vendor dependency in the architecture. Workflows define what should run, not how to wire up cloud infrastructure around them.
Apache Airflow, fully managed inside your AWS account. MWAA handles scheduler maintenance, worker scaling, and Airflow upgrades. DAGs are stored in S3, workers run in your VPC, and access is controlled through IAM. The AWS integration runs deep, and so does the dependency.
MWAA is Amazon's fully managed Airflow service—there is no local or self-hosted install option. This comparison reflects what's required to provision an MWAA environment: an S3 bucket with versioning enabled, a VPC with security groups, an IAM execution role, and environment provisioning that takes 20–30 minutes on its own.
curl -o docker-compose.yml \https://raw.githubusercontent.com/kestra-io/kestra/develop/docker-compose.ymldocker compose up
# Open localhost:8080# Pick a Blueprint, run it. Done.Download the Docker Compose file, spin it up, and you're ready (database and config included). Open the UI, pick a Blueprint, run it. No cloud account, no VPC, no S3 bucket, no IAM role.
# Step 1: Create S3 bucket with versioningaws s3api create-bucket --bucket my-mwaa-dags --region us-east-1aws s3api put-bucket-versioning \ --bucket my-mwaa-dags \ --versioning-configuration Status=Enabled
# Step 2: Configure VPC, security groups, IAM role...# Step 3: Create MWAA environment (20-30 min to provision)# Step 4: Upload DAGs to S3aws s3 cp dags/ s3://my-mwaa-dags/dags/ --recursive
# Now write your DAGs in Python...Create an S3 bucket with versioning enabled, configure a VPC with security groups, create an IAM execution role, provision the MWAA environment (20-30 min), then upload your DAG files to S3 and wait for the scheduler to pick them up.
YAML is readable on day 1. Our docs are embedded in the UI for easy reference, the AI Copilot writes workflows for you, or start with our library of Blueprints. No Python knowledge required to understand or modify a workflow.
DAG files are Python scripts stored in S3. To update a workflow, push a new file to S3 and wait for the scheduler to detect the change. Operators, dependencies, and scheduling are all Python. Non-Python contributors can view runs in the Airflow UI but cannot author or modify workflows.
Orchestrate across data pipelines, infrastructure operations, business processes, and AI workflows in one unified platform. Event-driven at its core, with native triggers for S3, webhooks, Kafka, and message queues. Runs on any cloud or on-prem.
A managed Airflow environment inside your AWS account. Excellent for Python-based data pipelines with deep AWS service integration. Infrastructure automation and business process workflows are possible via Python operators but are not first-class use cases. Runs exclusively on AWS.
| | | |
|---|---|---|
| Workflow definition | Declarative YAML | Python DAGs stored in S3 |
| Languages supported | Any (Python, SQL, R, Bash, Go, Node.js) | Python-first (Bash/SQL via operators) |
| Cloud dependency | Runs anywhere (AWS, GCP, Azure, on-prem) | AWS only |
| Setup requirements | Docker Compose (two commands) | S3, VPC, IAM role, MWAA environment (hours) |
| Architecture | Event-driven at core | Schedule-first (Airflow under the hood) |
| Airflow version control | Your upgrade timeline | AWS controls Airflow version availability |
| Multi-tenancy | Namespace isolation + RBAC out-of-box | IAM-based access control, separate environments for strong isolation |
| Self-service for non-engineers | Kestra Apps | Airflow UI is observability-focused, no self-service authoring |
| Infrastructure automation | Native support | Possible via Python operators and AWS SDK, not a primary use case |
We switched from Airflow because we want engineers solving problems, not coding orchestration. Kestra delivers end-to-end automation with the robustness we need at our scale. Few companies operate at this level, especially in AI/ML.
Kestra runs in two commands on Docker Compose. No S3 bucket to configure, no VPC to provision, no IAM role to get right before you can write a workflow. Teams get a production-shaped environment locally in under 5 minutes, with the same deployment model scaling to Kubernetes in production.
Kestra runs on AWS, GCP, Azure, or your own data center. When your data lives in multiple clouds or your team deploys some workloads on-prem, a single orchestrator that connects everywhere is simpler than maintaining separate tooling per environment. Native plugins for every major cloud give your team one control plane regardless of where workloads run.
Kestra upgrades are on your schedule: pull a new image, test in staging, and ship when your team is ready. No waiting for a managed service to certify a release or schedule a maintenance window. Roll back instantly if something unexpected surfaces.
Find answers to your questions right here, and don't hesitate to Contact Us if you couldn't find what you're looking for.
See how Kestra can simplify your data pipelines—and run them on any cloud without the AWS lock-in.