> ## Documentation Index > Fetch the complete documentation index at: https://docs.elitea.ai/llms.txt > Use this file to discover all available pages before exploring further. # Pipelines Overview > Welcome to the Pipeline Building Blocks guide! This comprehensive resource will help you understand and master ELITEA Pipelines—powerful visual workflows that a ## What are ELITEA Pipelines? ELITEA Pipelines are customizable, automated workflows that help you streamline repetitive tasks and complex processes. Think of a pipeline as a **step-by-step recipe** that guides the system through a series of actions to achieve a specific goal. Unlike traditional automation tools that require programming knowledge, ELITEA Pipelines can be created visually using a drag-and-drop interface, making powerful automation accessible to everyone—from business analysts to project managers. Pipeline Overview ### Key Characteristics **Visual Design** : Create workflows by connecting nodes in a flow diagram—no coding required. The Flow Editor provides an intuitive interface for designing even the most complex processes. **Flexible Automation** : Handle everything from simple linear tasks to sophisticated multi-step processes with conditional logic, loops, and integrations with external tools. **Smart Integration** : Connect seamlessly with toolkits, external services like Jira, GitHub, Confluence, and use AI agents to enhance your workflows. **Data-Driven** : Pipelines maintain their own memory (called "state") to track information as they execute, ensuring data flows smoothly between steps. ## How Do Pipelines Work? A pipeline operates by executing a sequence of **nodes**, where each node represents a specific action or decision. Here's how it works: 1. **Start at the Entry Point**: Every pipeline begins at a designated starting node 2. **Execute Nodes Sequentially**: The pipeline moves from one node to the next, performing actions and making decisions 3. **Manage State**: Information collected during execution is stored in the pipeline's state and can be used in subsequent steps 4. **Follow Connections**: Nodes are linked together, with the pipeline following these connections based on conditions or simple transitions 5. **Complete or Loop**: The pipeline ends when it reaches an END transition or continues looping based on your design ## Real-World Use Cases Pipelines excel at automating workflows that involve multiple steps, decisions, and integrations. Here are some practical examples: ### User Story Creation Workflow Automate the entire process of creating user stories in Jira: * Gather requirements from users through conversational prompts * Extract relevant information from Confluence pages or documentation * Enhance the story with AI-generated acceptance criteria * Get user approval before publishing * Automatically create the formatted user story in Jira ### Manual Test Case Generation Transform requirements into comprehensive test cases: * Analyze user stories or requirement documents * Generate test scenarios with step-by-step instructions * Create expected results for each test case * Add test cases to TestRail or Jira with proper categorization * Link test cases back to original requirements ### Other Common Scenarios * **Code Documentation**: Automatically document code repositories by analyzing files and generating comprehensive documentation * **Data Processing**: Extract, transform, and load data from multiple sources with validation and error handling * **Approval Workflows**: Route requests through multiple approvers with conditional logic based on criteria * **Report Generation**: Collect data from various systems, analyze it, and create formatted reports ## Pipelines vs Agents: When to Use Which? Both Pipelines and Agents are powerful automation tools in ELITEA, but they serve different purposes: ### Use Pipelines When: ✔️ **Structured Workflows**: You have a clear, defined sequence of steps that need to happen in order ✔️ **Multiple Integrations**: Your workflow requires connecting to several external tools or services (Jira, GitHub, Confluence, etc.) ✔️ **Conditional Logic**: You need to make decisions and branch the workflow based on specific conditions ✔️ **Data Processing**: You're handling data transformation, validation, or aggregation across multiple steps ✔️ **Orchestration**: You need to coordinate multiple agents, tools, or sub-processes as part of a larger workflow ### Use Agents When: ✔️ **Conversational Interactions**: You want a more free-form, chat-based interaction with users ✔️ **Single-Purpose Tasks**: The task is focused and doesn't require complex branching or orchestration ✔️ **Exploratory Work**: Users need to ask questions and explore information dynamically ✔️ **Simple Automation**: The workflow is straightforward without complex conditional logic Pipelines can include Agents as nodes! Use Agents for specific tasks within a larger Pipeline workflow to get the best of both approaches. ## Pipeline Building Blocks Understanding pipelines requires familiarity with their core components. Each building block serves a specific purpose in creating effective workflows: | Building Block | Description | Purpose | Key Topics | | ------------------------------------- | --------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------- | | **[States](./states)** | Pipeline's memory system for storing and managing data | Track information collected during execution and pass data between nodes | State variables, data types, input vs messages, state initialization | | **[Nodes](./nodes/overview)** | Individual actions or steps in your pipeline workflow | Perform specific tasks like interacting with users, calling services, making decisions, or processing data | 11 different node types organized into 4 categories (Interaction, Execution, Control Flow, Utility) | | **[Connections](./nodes-connectors)** | Links between nodes that control execution flow | Define how the pipeline moves from one step to the next, including conditional branching | Simple transitions, conditional edges, decision edges, router outputs, data mapping | | **[Entry Point](./entry-point)** | Designated starting node of your pipeline | Define where execution begins when the pipeline is triggered | Entry point syntax, validation rules, using different node types as starting points | | **[Flow Editor](./flow-editor)** | Visual interface for designing pipelines graphically | Build complex workflows using drag-and-drop without writing code | Creating nodes visually, connecting nodes, configuring parameters, debugging | | **[YAML Configuration](./yaml)** | Text-based format for defining pipelines programmatically | Enable advanced customization, version control, and programmatic pipeline generation | YAML schema, syntax rules, node configuration, complete examples | *** ## Getting Started with Pipeline Building Blocks Ready to dive deeper? We recommend exploring the building blocks in this order: 1. **Start with [States](./states)** - Understand how pipelines store and manage data 2. **Explore [Nodes](./nodes/overview)** - Learn about the different actions and components available 3. **Master [Connections](./nodes-connectors)** - Discover how to control the flow between nodes 4. **Understand [Entry Point](./entry-point)** - Define where your pipeline begins 5. **Choose Your Interface**: * [Flow Editor](./flow-editor) for visual design * [YAML Configuration](./yaml) for programmatic control 6. **Reference [Appendix - Comparison Tables](./appendix-comparison-tables)** - Quick comparisons of node types and features Before building pipelines, make sure you've completed the [Quick Start Guide](../../getting-started/chat-quick-start) and familiarized yourself with [Creating a Pipeline](../chat-conversations/how-to-create-and-edit-pipelines-from-canvas). ## Next Steps ### Learn by Doing * **[Create Your First Pipeline](../chat-conversations/how-to-create-and-edit-pipelines-from-canvas)**: Step-by-step tutorial for beginners * **[Pipelines Menu Guide](../../menus/pipelines)**: Complete reference for the Pipelines interface ### Explore Advanced Topics * **[Agent Integration](./nodes/interaction-nodes#agent-node)**: Incorporate AI agents into pipelines ### Get Inspired * **[Pipeline Examples](../../menus/pipelines#overview)**: Real-world pipeline implementations * **[Community Pipelines](https://nexus.elitea.ai/alita_ui/agents/latest)**: Discover public pipelines created by the community *** Ready to build powerful automated workflows? Start with **[States](./states)** to understand the foundation of pipeline data management!