Skip to main content

Overview

Bud AI Foundry Pipelines enable you to automate complex workflows through visual, directed acyclic graphs (DAGs). Build workflows that orchestrate deployments, manage clusters, and coordinate multi-step processes - all from an intuitive drag-and-drop editor. Pipeline Editor

Key Features

Visual Pipeline Editor Build workflows with drag-and-drop. Connect actions, configure parameters, and visualize execution flow - no code required. Event-Driven Execution Trigger pipelines automatically based on platform events like deployments, cluster changes, or custom triggers. Complete Observability Track every execution with detailed logs, status monitoring, and execution history stored in PostgreSQL. Built-in Reliability Automatic retries, conditional branching, and error handling ensure your workflows complete successfully.

Available Actions

Pipelines provide actions across multiple categories:
CategoryDescriptionKey Actions
Control FlowLogic and routingLog, Conditional Branch, Transform, Aggregate, Wait Until
DeploymentModel deployment operationsDeploy Model, Scale Deployment, Configure Rate Limiting
Model OperationsModel managementAdd Model, Add Cloud Model, Benchmark Model
ClusterInfrastructure operationsCluster Health Check
IntegrationExternal servicesHTTP Request, Send Notification
All actions support custom parameters, automatic retries, and timeout configuration. Use the visual editor to discover and configure actions.

How It Works

  1. Create - Build your pipeline in the visual editor
  2. Configure - Set up triggers (manual, scheduled, or event-based)
  3. Execute - Run pipelines on-demand or automatically
  4. Monitor - Track execution status and view detailed logs

Use Cases

Deployment Automation

Orchestrate model deployments with validation and rollback

Cluster Management

Automate health checks and maintenance workflows

Scheduled Jobs

Run recurring tasks on a schedule

Event Processing

React to platform events automatically

Getting Started