​
What can you do with it?

Search and retrieve information from your vector-enabled knowledge base collections using semantic search. Perfect for finding relevant content across large document sets, generating AI-powered answers from your data, and filtering results by metadata attributes. Ideal for documentation search, knowledge management, and content discovery with natural language queries. Get raw search results from your documents with relevance scores. Best when you need to see the actual content chunks and metadata without AI interpretation.

​
2. Semantic Search with AI Answer (RAG)

Get the same search results but with an AI-generated answer that synthesizes the information. Best when you want a direct answer to your question based on the found documents.

​
How to use it?

​
Basic Command Structure

/knowledgebase-collection [search-query]

​
Parameters

Required:
  • search-query - Natural language question or search terms
Optional:
  • action - Operation type: “ask” (default - generates AI answer) or “search” (returns raw results)
  • filters - Attribute filters to narrow search results
  • max-results - Number of results to return (1-50, default: 10)
  • model - AI model for answer generation (ask only, default: “gpt-4o-mini”)

​
Response Format

The command returns: 
{
  "vectorStoreId": "vector store identifier",
  "response": "AI-generated answer (ask only)",
  "data": [{
    "file_id": "file identifier",
    "filename": "document name",
    "score": "relevance score",
    "content": [{"type": "text", "text": "content snippet"}]
  }]
}

​
Examples

/knowledgebase-collection
search-query: customer data formats
action: search
Get raw search results about customer data formats from your knowledge base.

​
Semantic Search with Filters

/knowledgebase-collection
search-query: customer data formats
action: search
max-results: 20
filters: product_area = "crm"
Search for customer data information filtered to CRM-related documents only.

​
AI Answer Generation

/knowledgebase-collection
search-query: what are expense policies
Get an AI-generated answer synthesizing information about expense policies from your documents.

​
Notes

Use filters to narrow search results to specific documents or sections, reducing noise and improving relevance. Filters let you target exact subsets of your knowledge base based on metadata attributes. Available filter operators:
  • = equals
  • != not equals
  • > greater than
  • >= greater than or equals
  • < less than
  • <= less than or equals
  • AND combine multiple conditions (all must match)
  • OR combine multiple conditions (any can match)
Supported value types: strings (use quotes), numbers, booleans. Example: department = "engineering" AND last_updated > 2024