Download for macOS today, or join the waitlist for Windows and Linux.
Workflows / States

Parallel Workflows

Parallel states enable workflows to execute multiple branches at the same time, dramatically improving performance and modeling real-world processes that do not need to run sequentially. This collection of workflows showcases how Parallel states split execution into independent paths, each with its own set of states, and then synchronize the results into a single output. You will see how to run validations, enrichments, and external service calls concurrently, reducing overall latency while maintaining clear orchestration logic. These examples also highlight how to handle branch-level errors, merge outputs, and design workflows that remain resilient even when individual parallel paths fail or behave differently.

31 workflows

Bulk Booking Processing

Freight Booking

This workflow processes an array of freight bookings concurrently using a Map state with a maximum concurrency of 3. Each booking is independently subjected to carrier capacity checks and hazmat screening in parallel branches before being individually confirmed or cancelled.

Booking Amendment

Freight Booking

This workflow handles the end-to-end amendment of a previously confirmed freight booking. It validates the amendment request, enforces a mandatory processing hold using a Wait state, then runs concurrent carrier capacity re-check and hazmat re-verification using a Parallel state before issuing a confirmed amendment or cancellation.

Rag Ingestion Pipeline

RAG Knowledge Base

End-to-end RAG knowledge base ingestion pipeline covering document intake, text extraction and chunking, embedding generation, vector store indexing, retrieval quality validation, catalog update, and stakeholder notification.

Knowledge Base Refresh Escalation

RAG Knowledge Base

Escalation workflow for knowledge base refresh failures or stale content alerts. Holds for review, runs parallel re-extraction and re-indexing, then routes to resolution or manual intervention.

Bulk Document Ingestion

RAG Knowledge Base

Batch ingestion of multiple documents concurrently using a Map state. Supports scheduled full-corpus re-ingestion and large document library imports.

Model Training Pipeline

Model Training Pipeline

End-to-end MLOps pipeline covering data ingestion, feature preprocessing, SageMaker training job launch, model evaluation, registry registration, endpoint deployment, and stakeholder notification.

Model Drift Escalation

Model Training Pipeline

Escalation workflow for models flagged with performance drift. Applies a review hold, runs parallel re-evaluation and retraining, then routes to endpoint rollback or redeployment.

Bulk Model Retraining

Model Training Pipeline

Batch retraining of multiple models concurrently using a Map state. Supports scheduled monthly retraining runs across all production models.

Content Moderation Pipeline

Content Moderation

End-to-end AI content moderation pipeline covering content ingestion, multi-label classification, human review routing, moderation action enforcement, classifier feedback, and outcome notification.
Free Trial