Skip to main content
Upload your documents—PDFs, Word files, or text—and Wassist creates an agent that can intelligently search and answer questions about your content.

​
When to Use This Method

This approach is ideal when you have:
  • Product manuals or documentation
  • Training materials or guides
  • Research papers or reports
  • Internal knowledge bases
  • Any structured content in document form

​
How It Works

  1. You upload documents to Wassist
  2. Content is processed and indexed using AI embeddings
  3. When users ask questions, the agent searches your documents
  4. Relevant passages are used to generate accurate answers

​
Supported File Types

FormatExtensionNotes
PDF.pdfText-based PDFs work best; scanned documents may have reduced accuracy
Word.docxFull support for formatted documents
Text.txtPlain text files
Markdown.mdIncluding formatting
Maximum file size: 50MB per document. For larger files, consider splitting into smaller sections.

​
Step-by-Step Guide

1

Navigate to New Agent

From your dashboard, click New Agent.
2

Select 'Upload Knowledge Base'

Choose the Upload Knowledge Base option.
3

Name Your Agent

Give your agent a descriptive name that reflects the content it will know about.Examples:
  • “Product Documentation Assistant”
  • “HR Policy Helper”
  • “Research Paper Guide”
4

Upload Documents

Drag and drop your files or click to browse. You can upload multiple files at once.
For best results, organize related content into separate documents rather than one massive file.
5

Wait for Processing

Each document goes through:
  • Extraction — Text is extracted from the file
  • Chunking — Content is split into searchable sections
  • Embedding — AI creates semantic representations
Status indicators show progress for each file.
6

Customize and Test

Once processing completes:
  • Review the generated system prompt
  • Adjust the welcome message
  • Test with sample questions

​
Adding Documents to an Existing Agent

You can add more documents to any agent at any time:
  1. Open your agent in the editor
  2. Go to the Documents section
  3. Click Upload Documents
  4. Select files to add
New documents are processed and added to the existing knowledge base.

​
Document Processing Status

StatusDescription
PendingDocument is queued for processing
ProcessingContent is being extracted and indexed
ReadyDocument is fully indexed and searchable
ErrorProcessing failed (hover for details)
If a document fails, try:
  • Re-uploading the file
  • Converting to a different format (e.g., PDF → Word)
  • Extracting text manually to a .txt file

​
Best Practices

Keep related content in the same document or clearly labeled sections. This helps the AI understand context.Good: Separate PDFs for “Returns Policy”, “Shipping Info”, “Product Specs”Less effective: One giant document with everything mixed together
Make sure documents include enough context. If a document references “the product” without naming it, the AI may struggle.Good: “The XR-5000 requires a 120V power supply…”Less effective: “The product requires a 120V power supply…”
Well-structured documents with headings and sections are easier for the AI to navigate and cite.
Outdated documents lead to incorrect answers. Remove or update documents when information changes.

​
Customizing Knowledge Retrieval

In your agent’s system prompt, you can guide how documents are used: 
System Prompt Addition:

When answering questions:
- Always cite the source document name
- If you can't find the answer in the documents, say so clearly
- For complex topics, offer to explain in more detail
- If information seems outdated, recommend checking the official website

​
Managing Documents

​
Viewing Documents

Go to your agent’s Documents tab to see:
  • All uploaded documents
  • Processing status
  • File sizes and upload dates

​
Removing Documents

To remove a document:
  1. Click the document in the list
  2. Click Delete
  3. Confirm removal
The document is immediately removed from the knowledge base.

​
Replacing Documents

To update a document:
  1. Delete the old version
  2. Upload the new version
  3. Wait for processing to complete
Use clear version numbers in file names (e.g., product-manual-v2.pdf) to track which version is active.

​
SDK Alternative

Create a knowledge base agent programmatically: 
// Prepare files (in a browser environment)
const files = [
  document.getElementById('file-input').files[0],
  // ... more files
];

const agent = await client.onboarding.createFromKnowledgeBase({
  name: 'Product Documentation Assistant',
  files: files,
});

​
What’s Next

Add More Knowledge

Learn about document management.

Connect Tools

Combine documents with live data from APIs.

Deploy to WhatsApp

Make your knowledge agent live.