Pegasus AI Studio
Design, execute, and monitor large-scale scientific workflows at any scale — with an AI assistant that manages the complexity so you can focus on the science.
A unified platform for scientific workflow execution
Research institutions worldwide have depended on Pegasus WMS for over two decades to orchestrate large-scale scientific pipelines — from gravitational wave detection and genome sequencing to climate modeling and particle physics.
PegasusAI Studio brings that power to the browser: describe your pipeline in natural language, let the AI generate and configure it, then execute it across hundreds or thousands of cores with a single click.
From research goal to running pipeline in three steps
Describe your scientific goal in plain language. The AI translates your intent into a fully configured Pegasus WMS workflow — complete with scripts, dependencies, and execution settings.
Your workflow is compiled and dispatched automatically — running on your local cluster, your institution's computing infrastructure, or distributed across global shared pools.
Track every job as it executes across your pipeline. The AI surfaces failures, diagnoses root causes, and recommends corrective actions before issues cascade.
Capabilities built for research at scale
AI Assistant
Natural language workflow authoring
- Compose pipelines by describing your scientific goals
- Diagnose failures with AI-powered root-cause analysis
- Auto-generate execution scripts and configuration files
- Expert Pegasus WMS guidance without reading documentation
Live Dashboard
Complete pipeline observability
- Interactive DAG visualization of your full workflow graph
- Per-job execution status updated in real time
- One-click access to logs, stderr output, and exit codes
- Proactive alerts when jobs stall, fail, or exceed resource limits
Notebooks
Integrated computational environment
- Browser-based Jupyter notebooks with zero configuration
- Immediate access to your workflow inputs, outputs, and data
- Python, R, and Julia kernels available on demand
- AI assistant embedded alongside your code for in-context help
Compute Power
Elastic infrastructure at any scale
- Execute workflows on your local or institutional computing cluster
- Seamlessly burst to OSPool's 19,000+ cores across global sites
- Real-time resource utilization and quota visibility
- Threshold alerts before capacity limits are reached
A full walkthrough of the interface
Project setup, workflow visualization, and live job monitoring — in one take.
Studio Orientation
Workflow Tutorial
Start your first pipeline today
Researchers can request access at studio.pegasus-ai.org. Sign-in uses CILogon with institutional credentials or another supported provider.
Powered by Pegasus WMS · Apache 2.0 · Open Source