This guide provides comprehensive documentation for the Catalyzed platform's Pipeline API, enabling advanced data processing workflows through configurable block-based pipelines.

Overview

The Pipeline API provides a powerful framework for creating and executing data processing workflows. Pipelines are composed of reusable blocks that can be connected to form complex processing chains, supporting everything from simple data transformations to sophisticated AI/ML operations.

Key Features:

• Block-Based Architecture - Compose pipelines from 40+ pre-built processing blocks
• YAML Configuration - Define pipelines using human-readable YAML syntax
• Type-Safe Execution - Automatic type inference and validation across blocks
• Async Processing - High-performance parallel block execution
• AI/ML Integration - Built-in blocks for LLMs, embeddings, and semantic operations
• Progress Tracking - Real-time execution monitoring with detailed progress updates

For authentication, request formatting, pagination, rate limits, and other common API patterns, please refer to our comprehensive REST API documentation.

Base URL: https://platform-api.us.catalyzed.ai

Core Concepts

Pipeline Lifecycle

1. Create Pipeline - Establish pipeline metadata and permissions
2. Define Configuration - Create configuration version with YAML pipeline definition
3. Execute Pipeline - Start pipeline execution with input data
4. Monitor Progress - Track execution progress and block status
5. Retrieve Output - Access results from completed pipeline