Skip to main content

Knowledge Base

The Knowledge Base lets you upload documents that your AI agents can reference during calls. When a customer asks a question, your agent can search your knowledge base for accurate, up-to-date answers.

How It Works

  1. You create Collections to organize your documents
  2. You upload documents to collections
  3. Documents are automatically processed and indexed
  4. AI agents search the knowledge base during calls
  5. Agents use retrieved information to answer questions
Knowledge retrieval uses semantic search, meaning the AI finds relevant information based on meaning, not just keyword matching.

Creating a Collection

Collections help you organize related documents:
1

Navigate to Knowledge

Go to Dashboard → Knowledge and click New Collection.
2

Name Your Collection

Give it a descriptive name (e.g., “Product FAQs”, “Pricing Information”).
3

Add Description

Describe what this collection contains to help you stay organized.
4

Configure Settings

  • Enable for Campaigns: Allow AI agents to use this collection
  • Public/Private: Control visibility
5

Save

Click Create to save the collection.

Collection Categories

Organize collections by type:
CategoryUse For
Product SpecificationProduct details, features, specs
FAQFrequently asked questions
LegalTerms, policies, compliance info
Training MaterialInternal guides and procedures
PricingPricing sheets, discounts, packages
GeneralMiscellaneous documents

Uploading Documents

1

Open a Collection

Click on a collection to view its contents.
2

Click Upload

Click the Upload Documents button.
3

Select Files

Choose one or more files from your computer.
4

Wait for Processing

Documents are automatically processed in the background.

Supported Formats

FormatExtensionMax Size
PDF.pdf50 MB
Word.doc, .docx50 MB
Text.txt50 MB
Markdown.md50 MB
For best results, use well-structured documents with clear headings and organized content.

Document Processing

When you upload a document, EolasFlow automatically:
  1. Extracts text from the document
  2. Chunks content into smaller pieces for better retrieval
  3. Creates embeddings (vector representations for semantic search)
  4. Indexes the content for fast retrieval

Processing Status

StatusMeaning
PendingWaiting to be processed
ProcessingCurrently being analyzed
CompletedReady for use
FailedProcessing error (check format)
Processing typically completes within a few seconds for most documents.

Using Knowledge in Campaigns

To give an AI agent access to your knowledge:
  1. Go to Dashboard → Campaigns
  2. Edit or create a campaign
  3. In Knowledge Base settings, select the collections to include
  4. Save the campaign
During calls, the agent will automatically search these collections when relevant questions arise.

Managing Collections

Edit a Collection

  1. Click on the collection
  2. Click Edit
  3. Update name, description, or settings
  4. Save changes

Delete a Collection

  1. Open the collection
  2. Click Delete
  3. Confirm deletion
Deleting a collection removes all documents and their embeddings. This cannot be undone.

View Collection Stats

Each collection shows:
  • Document Count - Number of uploaded files
  • Total Size - Storage used
  • Views - How often accessed
  • Agent References - Times used by AI agents

Document Management

View Document Details

Click on any document to see:
  • File name and size
  • Upload date
  • Processing status
  • Number of chunks created

Delete a Document

  1. Open the collection
  2. Find the document
  3. Click the delete icon
  4. Confirm deletion

Best Practices

Keep Documents Current: Regularly update your knowledge base with the latest information.
Organize Logically: Group related documents into collections for easier management.
Use Clear Language: Write documents in clear, conversational language that AI can easily reference.
Test Retrieval: Ask your AI agent questions to verify it finds the right information.
Include FAQs: Upload FAQ documents for common questions your agents should handle.

Example Use Cases

IndustryKnowledge Collections
SaaSProduct docs, pricing, integrations, FAQs
HealthcareServices, insurance info, appointment policies
Real EstateProperty listings, financing options, process guides
E-commerceProduct catalogs, shipping policies, returns
Your AI agent uses semantic search, which means:
  • Meaning-based: Finds content by meaning, not just keywords
  • Context-aware: Understands the context of questions
  • Fuzzy matching: Handles synonyms and related terms
For example, if a customer asks “What’s the price?”, the AI can find content about “cost”, “pricing”, “fees”, or “rates”.

Next Steps