Knowledge Bases
Knowledge Base actions in Lleverage allow you to create, manage, and utilize collections of information that enhance your workflows with specialized content. These actions make it easy to store documents, add context, and reference specific information throughout your automations.
Key Features
Document storage for various file types (PDFs, Word docs, CSV, JSON, XML, HTML)
Text content management for quick manual additions
Contextual knowledge that enhances AI capabilities
Searchable information accessible throughout workflows
Advanced configuration options to optimize information retrieval
Shared resources that can be used across multiple workflows
How to Add Knowledge Base Actions
Open the Add Action menu using one of three methods:
Click the "Add Action" button in the top left corner
Click on a connection circle on an existing action card
Click and drag from one node to create a connection
Find Knowledge Base actions in:
The "All" tab
The "Tools" tab
The "Storing and Using Information" subcategory
Choose from the available Knowledge Base actions:
Add Knowledge: Store information in a Knowledge Base
Search Knowledge Base: Query and retrieve information
Click on your chosen action to add it to your canvas
💡 Tip: These actions help you manage information within your workflows, making knowledge accessible when and where it's needed.
How to Configure Add Knowledge Action
Setting Up the Basic Fields
When you select the Add Knowledge action, you'll see three main fields:
Content (optional): The information you want to store
Title (optional): A descriptive name for the content
Knowledge Base: The destination folder (defaults to "knowledge base")
Selecting or Creating a Knowledge Base
Click on the Knowledge Base dropdown
Select an existing Knowledge Base if available
Or click "Create Knowledge Base" to make a new one
If Creating a New Knowledge Base:
A new Knowledge Base will be created (titled "Untitled" or named after your workflow)
Click the edit icon (cog) to customize your Knowledge Base
The Edit Knowledge Base modal will open with advanced options
💡 Tip: Adding content to a Knowledge Base makes it available for your workflow to reference, similar to giving your workflow access to a specialized document library.

How to Configure Advanced Knowledge Base Settings
Primary Content Type Configuration
This setting determines how documents are processed and chunked for optimal retrieval:
Auto: Automatically detects the best content type and chunking strategy
Structured Data: For CSV, JSON, XML files with custom chunking to preserve data integrity
Mixed/Web Content: For webpages and HTML files with basic chunking (300 characters)
Structured Documents: For PDFs and Word docs, chunks by headings and sections (250 characters)
Narrative Content: For long-form text and articles with similarity-based chunking (350 characters)
Transactional Documents: For forms and receipts with no chunking (keeps documents intact)
Conversational Content: For chat logs and transcripts with similarity-based chunking (100 characters)
Page-bound Content: For presentations and page-based documents, chunks by page boundaries
Language Settings
Default setting is "Multilingual"
This determines the embedding model used to vectorize your documents
Optimizes processing based on the language of your content
Adding Content to Your Knowledge Base
Upload Files: Add documents directly from your computer
Write Something: Opens a text input to create content from scratch
After adding content, you can:
View uploaded files with checkmarks indicating successful upload
Delete items as needed
Rename your Knowledge Base
⚠️ Important: Choose the appropriate Content Type based on the kind of information you're storing to optimize how the system processes and retrieves it.
How to Configure Search Knowledge Base Action
Basic Setup
Add the "Search Knowledge Base" action to your workflow
Configure the required Query field with your search terms
Select which Knowledge Base to search from the dropdown
Search Configuration Options
Match Search Results: Set between 1-50 results (maximum 50, minimum 1)
Search Fuzziness: Scale from 0-100
0: Keyword-based search (exact matches)
100: Semantic search (meaning-based matches)
Mid-range: Combination of both approaches
Using Search Results
Connect the search results to other actions in your workflow
Typically used with AI actions that need contextual information
Results can be referenced using variables in subsequent workflow steps

How Knowledge Base Actions Work in Your Workflow
Use Case: Building a Knowledge Repository
Add the "Add Knowledge" action to your workflow
Select or create a Knowledge Base
Configure content to be added (directly or via workflow variables)
When the workflow runs, information is automatically stored
Useful for:
Capturing important information during workflow execution
Building a Knowledge Base over time with new data
Creating a permanent record of workflow results
Use Case: Retrieving Contextual Information
Add the "Search Knowledge Base" action to your workflow
Select which Knowledge Base to search
Configure the search query (can be dynamic based on workflow variables)
Connect search results to AI actions for enhanced responses
Useful for:
Providing context to AI prompts
Retrieving specialized information for decision-making
Creating responses based on your organization's knowledge
💡 Pro Tip: Combining both actions allows you to build workflows that continuously learn and improve by adding new information and then accessing it when needed.
Best Practices and Important Considerations
Knowledge Base Management
Knowledge Bases are shared resources across multiple workflows
You can add multiple documents to a single Knowledge Base
Files are displayed with checkmarks when successfully uploaded
Use the cog icon to access advanced configuration options
Content Optimization
Different content types work better with specific primary context settings
More specific search queries typically yield better results
Consider the chunking strategy based on your document types
Workflow Integration
Content added to Knowledge Bases enhances AI capabilities by providing specialized context
Search results can be used as variables in subsequent workflow steps
Knowledge Bases persist across workflow runs, building institutional knowledge
💡 Think of it this way: Knowledge Bases are like specialized libraries that your workflows can both contribute to and reference for better, more informed results.
Last updated
Was this helpful?