Pipelines

​

Overview

• Workflow Orchestration: Automate multi-step processes with conditional logic, loops, and parallel execution paths
• Visual Design: Design complex workflows using an intuitive drag-and-drop interface or advanced YAML configuration
• Service Integration: Connect multiple tools and services in a single workflow for seamless data flow
• Reusability: Create modular, reusable workflows that can be nested and combined for different use cases
• Scalability: Handle both simple linear processes and intricate multi-branch workflows with equal efficiency

1. Node-Based Design: Build workflows by connecting nodes representing different actions—LLM calls, agent interactions, toolkit operations, code execution, routers, and decision points
2. Visual or Code Configuration: Use the Flow Designer for visual workflow creation or switch to YAML Editor for advanced programmatic configuration
3. Execution & Integration: Connect toolkits, agents, and MCPs to extend capabilities; configure model settings and conversation flows for interactive pipelines

• Visual Workflow Design: Intuitive Flow Designer with drag-and-drop nodes for easy workflow creation and management
• Conditional Logic: Implement complex decision trees using Router and Decision nodes for dynamic workflow paths
• Multi-Service Integration: Connect external toolkits, agents, MCPs, and nested pipelines within a single workflow
• Flexible Configuration: Toggle between visual Flow Designer and YAML Editor for different levels of customization
• State Management: Track and modify workflow state throughout execution using State Modifier nodes
• Version Control: Create, manage, and compare different versions of pipelines for iterative development

​

Navigating the Pipelines Menu

​

Pipelines Dashboard

• Card View - Visual cards displaying pipeline name, description, and key information. Ideal for browsing and quick identification.
• Table View - Compact table format with columns for name, description, status, and metadata. Better for managing large numbers of pipelines.

• Search Bar - Quickly find pipelines by typing the pipeline name, description, or related keywords
• Filter by Tags - Filter pipelines by their assigned tags to find workflows in specific categories or with particular characteristics. Click tags to narrow down the list.

1. Locate the pipeline you want to pin
2. Click the pin icon (📌) on the pipeline card or in the table row
3. Pinned pipelines will appear at the top of the list, separated from unpinned ones
4. Click the pin icon again to unpin the pipeline

​

Creating a Pipeline

1. Click the + Create button located in the sidebar navigation.
2. In the General section, fill out the required fields:
   • Name: Enter a unique name for your pipeline (e.g., “Test Case Generator Pipeline”)
   • Description: Provide a clear description of the pipeline’s purpose (e.g., “Automatically generates comprehensive test cases from acceptance criteria and user stories”)
3. Optionally, add and configure:
   • Tags: Add tags by typing a tag name or selecting from pre-existing tags
   • Welcome Message: Define a welcome message to guide users interacting with the pipeline. (e.g., “Welcome! I can help you generate test cases based on your acceptance criteria. Simply provide your requirements and I’ll create comprehensive test scenarios for you.”)
   • Conversation Starters: Add predefined commands or prompts to initiate interactions with the pipeline. (e.g., “Generate test cases for the login feature” or “Create test scenarios for payment processing”)
4. In the Advanced section (optional):
   • Step Limit: Set the maximum number of tool execution steps the pipeline can perform in a single turn (range: 0-999, default: 25). This parameter controls how many iterations the pipeline can execute before stopping, preventing infinite loops and managing resource usage. A higher step limit allows more complex multi-step workflows, while a lower limit ensures faster execution for simpler tasks.
5. In the Editor Notes section (optional):
   • Notes: Add internal documentation for the current pipeline version, such as rollout reminders, implementation notes, or review comments. These notes are for editors only and are not sent to the LLM or used during execution.
6. Click Save to create the pipeline. This action creates what’s known as the “base” version of your pipeline.

• Left panel — the configuration form with all settings sections (General, Tools, Variables, Welcome Message, Conversation Starters, Advanced, Editor Notes, Information)
• Center panel — the Flow/YAML editor for visually designing or coding your pipeline workflow
• Right panel — an embedded live chat panel for testing the pipeline without leaving the editor

• Add Toolkits: Integrate external services and APIs to extend your pipeline’s capabilities
• Add Agents, MCPs, and nested Pipelines: Include other agents, Model Context Protocol servers, or other pipelines to extend functionality
• Design the Workflow: Build the pipeline flow using the visual Flow Designer or the YAML editor
• Select AI Model: Choose the appropriate language model and configure model settings
• Test Interactively: Use the embedded chat panel on the right to send messages and verify pipeline behaviour in real time
• View Run History: Click the clock icon (🕐) in the top bar of the chat panel to open the Run History view

When configuring Pipelines, you can further personalize their profiles by adding a custom image along with the Name and Description. This feature allows you to create a unique, visually distinct identity for each Pipeline, making them easier to recognize and manage.To add an image:

1. Click the Pen Icon next to the image placeholder. Clicking this icon will open the image upload interface.
2. Click the Upload a Custom Image icon to upload a custom image from your local system to personalize the Agent’s profile.
3. Use Default Images from a set of default images provided by the platform.

​

How to Select and Configure Tools

1. In the pipeline Configuration interface, navigate to the TOOLS section
2. Use the dedicated add buttons to attach resources. Each button opens its own searchable dropdown:
   • +Toolkit: Browse and select from available toolkits, or click “Create new” to create a new toolkit. See Toolkits.
   • +MCP: Browse and select Model Context Protocol servers, or click “Create new” to create a new MCP. See MCPs.
   • +Agent: Browse and select agents to enable your pipeline to delegate tasks or collaborate with specialized agents. See Agents.
   • +Pipeline:Browse other pipelines to create complex multi-step workflows by chaining pipelines together
3. Each dropdown supports inline search and paginated scrolling to locate resources quickly.

Note: Your changes are saved automatically when you add or remove resources.

​

EDITOR NOTES

• The content is not sent to the LLM, not included in chat or execution, and not used in monitoring

• Use this version for staging validation only.
• Pending toolkit credential rotation before production rollout.
• Draft variant created for testing alternative routing logic.

​

Designing Your Pipeline

​

State Management

• Pass data between nodes without explicit connections
• Store intermediate results for later use
• Implement conditional logic based on accumulated data
• Track workflow progress and decision outcomes

1. In the left panel, locate the State section
2. Click + Context to create a new state variable
3. Enter a Variable Name (use descriptive names like project_id, analysis_result, counter)
4. Set an default Value (optional) - can be text, numbers, or JSON objects
5. Click Save to apply your changes

• State Modifier Nodes: Use these specialized nodes to read from and write to state variables during pipeline execution
• Reference in Nodes: Access state variables in other nodes using the syntax {{state.variable_name}}
• Conditional Logic: Use state variables in Decision and Router nodes to control workflow branches
• Data Transformation: Pass state values between LLM, Code, and Toolkit nodes for complex processing

​

Flow Designer:

• Use Flow mode (selected via the Flow/YAML toggle in the editor panel) to visually design your pipeline by connecting various nodes, such as LLM, Agent, Toolkit, MCP, Code, Custom, Router, Decision, State Modifier, and Printer.
• Add new nodes by clicking the + icon and selecting the desired node type from the dropdown menu.
• Use the End node to define the completion of the pipeline.
• Drag and drop connections between nodes to establish the workflow’s logic and transitions.
• Zoom in or out and adjust the view for better navigation and management of complex workflows.

​

YAML Editor:

• Switch to YAML mode using the Flow/YAML toggle in the editor panel for code-based configuration.
• Define complex workflows, conditions, and logic that may not be easily achievable through the visual Flow Designer.
• Use the YAML editor to fine-tune node configurations, set advanced parameters, and integrate custom logic.
• Validate your YAML syntax to ensure the pipeline runs smoothly without errors.

• This is particularly useful for users who prefer a code-first approach, need to implement intricate logic, or want to replicate and modify existing pipelines efficiently.

​

How to Execute Pipeline

1. Configure the Pipeline: Ensure your pipeline is properly configured with clear instructions and any necessary tools.
2. Navigate to Pipeline: Access the pipeline’s interface by clicking on the pipeline from your pipelines list.
3. Select the AI Model: In the interface, choose the appropriate AI model (e.g., gpt-4o, gpt-5.1, etc.) from the model selection dropdown.
4. Adjust Model Settings (Optional): Click the Model Settings icon (⚙️) next to the model selector to fine-tune the response generation. The settings vary depending on the selected model: For Reasoning Models (e.g., GPT-5.1):
   • Reasoning - Controls the depth of logical thinking and problem-solving with three levels:
     • Low: Fast, surface-level reasoning with concise answers and minimal steps
     • Medium: Balanced reasoning with clear explanations and moderate multi-step thinking (default)
     • High: Deep, thorough reasoning with detailed step-by-step analysis (may be slower)
   For Standard Models (e.g., GPT-4o):
   • Creativity - Controls response randomness and creativity. Lower values produce more focused and deterministic outputs, while higher values generate more diverse and creative responses with five levels (1-5):
     • 1: Highly focused and deterministic outputs
     • 2: Mostly focused with slight variation
     • 3: Balanced between focus and creativity (default)
     • 4: More varied and creative responses
     • 5: Maximum creativity and diversity
   Max Completion Tokens Limits the maximum length of AI responses measured in tokens (roughly 4 characters per token).(All Models):
   • Auto (default): System automatically sets the token limit to 4096 tokens
   • Custom: Manually set a specific token limit for responses
     • When Custom is selected, you can enter a specific number of maximum tokens
     • The interface shows remaining tokens available after your specified limit
     • Setting too high a value will show an error if it exceeds the model’s maximum output tokens
   
5. Start Interaction: Begin your conversation by either:
   • Clicking on a Conversation Starter (if configured) to use a predefined prompt
   • Typing your question or command directly into the chat input field
   • Using simple commands like “Go”, “Start Generating”, “Execute”, or “Run it” followed by clicking the Send button

• Full Screen Mode: Increase the size of the output window for better visibility and focus. This mode can be activated to expand the output interface to the full screen.
• Continue the Dialogue: To keep the conversation going, simply type your next question or command in the chat box and click the Send icon.
• Copy the Output: Click the Copy to Clipboard icon to copy the generated text for use elsewhere.
• Regenerate Response: If the output isn’t satisfactory, click the Regenerate icon to prompt the Gen AI to produce a new response.
• Delete Output: To remove the current output from the chat, click the Delete icon.
• Purge Chat History: For a fresh start or to clear sensitive data, click the Clean icon to erase the chat history.
• Like or Dislike the Output:
  • Click the Like icon if the output meets your expectations.
  • Click the Dislike icon if the output is unsatisfactory. Upon disliking, you will have the option to leave a comment explaining why the output did not meet your expectations. This feedback helps improve the system’s performance and relevance.
  

​

Managing Context Budget

• Real-time Token Tracking: Monitor token consumption as your pipeline processes requests
• Automatic Pruning: System removes older messages when context limit is reached
• Conversation Summarization: Generate summaries of conversation history to preserve context
• Message Preservation: Configure how many recent messages to always keep
• Flexible Strategies: Choose between different pruning strategies based on your needs

1. Ensure the context_manager secret is set to true in your project settings
2. Start executing your pipeline
3. The Context Budget widget appears above the chat interface
4. Click to expand and view detailed metrics and configuration options

​

Pipeline Runs

• Execution Tracking: Monitor the progress and status of each pipeline execution in real-time
• Input/Output Capture: Review the exact inputs provided and outputs generated for each run
• State Evolution: Track how state variables change throughout the execution
• Performance Metrics: Analyze execution time, token usage, and resource consumption
• Error Diagnostics: Identify where failures occur and access detailed error messages

​

Voice Capabilities

​

Voice Input

1. Click the microphone icon in the message input toolbar to start recording.
2. Speak your prompt. A live transcript appears in the input field as you talk — interim results update in real time.
3. Click the Stop (■) button to finish recording. The final transcript is committed and focus returns to the input field.

​

Text-to-Speech

1. Click the Read out (megaphone) button below a pipeline response to start playback.
2. The text is highlighted as it is read aloud.
3. A playback pill appears in the input area with Pause and Resume controls.
4. Click Pause to stop playback mid-sentence; click Resume to continue from where it left off.

​

Speaking Mode

1. Leave the message input field empty.
2. Click the voicewave icon (shown in place of the Send button when the input is empty) to enter Speaking Mode.
3. A voicewave pill with an Exit (✕) button replaces the send button, indicating Speaking Mode is active.
4. Speak your prompt. After 3 seconds of silence, your message is automatically sent.
5. The pipeline response streams in and is read aloud via Text-to-Speech.
6. Once the response finishes playing, recording starts again for the next turn.
7. Click the ✕ button in the voicewave pill to exit Speaking Mode at any time.

​

File Attachments

1. Open the pipeline and locate the left configuration panel.
2. Scroll to the INTERNAL TOOLS section.
3. Toggle the Attachments switch to the on position.
4. Click Save to apply the change.

• Paperclip icon (📎) — Click the paperclip button in the message input toolbar to open the system file picker and select one or more files. The tooltip shows how many attachment slots remain (“Attach files (N left)”) and is disabled once the maximum is reached.
• Drag and drop — Drag files from your operating system and drop them directly onto the chat input area.
• Clipboard paste — Copy files to your clipboard and press Ctrl+V while the chat input is focused to attach them instantly.

• Images are displayed as thumbnail previews in a grid layout.
• Other files (documents, code, data, archives) are displayed as filename chips showing the file name and type icon.

state:
  input_attachments:
    type: list
    default: []

​

Managing Pipeline Versions: Save, Create Versions, and Manage

• To save your work on a Pipeline for the first time, simply click the Save button. This action creates what’s known as the “base” version of your pipeline.
• You can continue to modify your pipeline and save the changes to the “base” version at any time by clicking the Save button again. If you wish to discard any changes made, you have the option to click the Discard button before saving.

​

How to Create New Versions:

1. Initiate a New Version: Start by clicking the Save As Version button.
2. Name Your Version: When saving your work, provide a version name that clearly identifies the iteration or changes made.
3. Click Save to confirm your entry.

• Delete: Remove this version of the Pipeline if it’s no longer needed.
• Execute: Run this specific version of the Pipeline to see how it performs.
• Navigate Versions: Use the Version dropdown list to switch between and select different versions of the Pipeline. This allows for easy comparison and management of various iterations.

​

Exporting and Importing Pipelines

1. Open the pipeline and select the version you want to export using the version selector.
2. Click the three-dot menu (⋮) in the toolbar.
3. Select Export. The file downloads automatically to your device.
   • If the selected version has no nested dependencies, a single .md file is downloaded.
   • If the selected version references other agents or pipelines as tools, the export is a .zip file containing all dependencies.

1. In the Pipelines dashboard, click the Import button in the toolbar.
2. Upload the .md or .zip file previously exported from ELITEA.
3. The Import Wizard displays entity cards for the main pipeline and any nested dependencies. Review configuration before confirming.
4. Click Import to complete. The pipeline will appear in the current project.

​

Forking Pipelines

1. Open the pipeline and select the version you want to fork using the version selector.
2. Click the three-dot menu (⋮) in the toolbar and select Fork.
3. In the Fork parameters wizard, choose a target project from the dropdown. No project is pre-selected; you must pick one manually.
4. Review the entity cards — the main pipeline card and any nested dependencies that will be forked automatically.
5. Click Fork. When complete, click Got it to navigate directly to the forked pipeline in the target project.

​

Viewing Pipeline History

1. Open any saved pipeline.
2. In the right panel (embedded chat area), click the clock icon (🕐) in the top bar. The tooltip reads “View run history”.
3. The pipeline page is replaced by the two-panel Run History display. Click the ✕ (Close) button in the top-left to return.

• Review past conversations: View complete chat histories from previous pipeline runs
• Track performance: Monitor execution duration across different versions
• Debug issues: Replay conversations to identify where problems occurred
• Compare versions: See how different pipeline versions performed with the same inputs
• Audit trail: Maintain records of all pipeline interactions for compliance purposes

• Left panel — sortable list of all past runs. Click any column header to sort:

  Column	Description
  Date	Timestamp of the run in dd-MM-yyyy, hh:mm a format
  Version	Pipeline version used for the run
  Duration	Total execution time of the run

  Click a row to load the conversation replay in the right panel. Each row also has a three-dot menu (⋮) with: Share link, Delete, and Restore conversation.
• Right panel — full conversation replay of the selected run.

​

INFORMATION

​

​

Best Practices

​

Troubleshooting

​

Support Contact