Hello World Example: Getting started with Kestra
Your first Kestra flow: take a user input, pass values between tasks, and log a personalized greeting.
New to Kestra? Start with these six core patterns.
Your first Kestra flow: take a user input, pass values between tasks, and log a personalized greeting.
An end-to-end extract, transform, load pipeline that pulls JSON from an API, cleans it in Python, and aggregates it with DuckDB.
Build a SQLite table, query it, and export the results to CSV and Excel for business stakeholders.
A human-in-the-loop vacation approval workflow that pairs a Slack approval request with an automated API submission in Kestra.
Build a Docker image, run it, and provision files with Terraform in one declarative Kestra pipeline.
Probe an API endpoint on a schedule and fire a Slack alert the moment it stops returning HTTP 200.
Community favorites: AI agents, Kubernetes, Terraform, Kafka, dbt, and Snowflake at work.
Classify every prompt with OpenAI, then route it to the best-fit LLM across Anthropic, Gemini, Perplexity, and OpenAI for quality and cost.
Order a Google Cloud VM through a Slack-notified approval gate, provision it with Terraform, and register it in inventory.
Plan Terraform, email a reviewer, pause for human approval, then apply the exact reviewed plan.
Provision a server fleet from a JSON manifest and register every machine as a typed VM asset with env, OS, and IP metadata.
A realtime Kafka trigger starts one execution per record, branches on the payload, and alerts Slack on anomalies.
Poll ServiceNow for new incidents on a schedule and fan out each one to a remediation subflow.
The newest blueprints in the catalog.
Detect Apache Pulsar subscriptions with no consumers that still pin a large backlog, then clear or remove them behind a dry run, an approval pause and a live re-read that fails the run.
Pause a purchase order for a finance manager's decision and return the verdict and comments as typed flow outputs.
Policy ratchet around pulumi preview that issues, alerts, and holds for approval before applying violating plans.
Stream Apache Pulsar messages straight into MySQL in real time, one row per event.
Recovery subflow that restores a flaky system to a healthy state so a parent self-healing loop can succeed on its next attempt.
Ingest JSON from Amazon S3, unify and curate it with DuckDB, and register every step as governed dataset assets with full lineage.
Find blueprints by domain: what part of your stack they touch.
LLM calls, RAG pipelines, and agentic workflows.
256+ blueprints BusinessApprovals, reporting, and cross-team automation.
339+ blueprints CloudAWS, GCP, and Azure services wired into one flow.
389+ blueprints CoreKestra-native mechanics: triggers, conditions, flow control.
248+ blueprints DataELT, warehouses, and dbt-driven transforms.
939+ blueprints InfrastructureProvisioning, patching, and configuration management.
1290+ blueprints
Copy a blueprint, adapt it to your stack, and ship your first flow in minutes.
Find answers to your questions right here, and don't hesitate to Contact Us if you couldn't find what you're looking for.