GraceIntroduction
Grace is an open source orchestration toolchain for running hybrid workflows across mainframe and cloud systems.
It brings declarative config, modular jobs, and unified logging to the world of z/OS. With Grace, you define, run, and monitor cross-platform workflows using a single YAML file. No brittle JCL chains. No polling SDSF.
Why Grace?
Mainframes power banking systems, government programs, transportation networks, and more. But automating against them is an exercise in frustration.
Grace simplifies orchestration by codifying mainframe jobs in a shared, declarative format. It gives teams a consistent way to define and run batch jobs, whether they're compiling COBOL, uploading to S3, or transforming data in Python.
Built for hybrid teams
In many orgs, mainframe and cloud teams operate in silos. Mainframe engineers manage legacy systems; data engineers extract and load that data downstream. Grace bridges the gap.
- Mainframe teams use Grace to define extract and batch jobs.
- Cloud teams run ETL, trigger analytics, and chain cloud-native steps.
- Platform teams operate Grace as the orchestration layer in between.
By working on a central control plane, teams gain joint ownership of the full pipeline - from JES to RDS, shell script to SQL.
What can you do with Grace?
If you've used Terraform, GitHub Actions, or Airflow, Grace will feel familiar.
With one workflow file, you can:
- Compile COBOL programs
- Run batch jobs with inputs and outputs
- Move datasets to cloud storage (e.g. S3)
- Trigger ETL pipelines and Python scripts
- Chain steps across z/OS, shell, and cloud platforms
- Monitor, retry, and log the whole operation from one place
Who is Grace for?
Grace is for teams automating around the mainframe.
- Mainframe engineers tired of brittle JCL chains
- Data and cloud engineers who need reliable, repeatable handoffs
- Platform teams standardizing job orchestration across systems
- Migration consultants building reproducible flows for legacy workloads
- Enterprises modernizing incrementally, not all at once
The name
Grace is named after Grace Hopper—pioneer of COBOL and early advocate for human-readable programming. She believed programming should be accessible and expressive. We're bringing that same spirit to infrastructure automation, across the most enduring systems in computing history.
Where to next?
Install Grace
Install the Grace CLI from a shell script, prebuilt binaries, or compile from source.
Configure your environment
Set up Zowe, authenticate with z/OS, and prepare your system to run workflows.
Hello, Grace
Run your first workflow with a sample COBOL job and dataset handoff.
CLI Reference
See command usage for building, validating, and submitting Grace workflows.