Semantic search in Knowledge Base

Performs semantic search using vector similarity in a specific Knowledge Base. Uses embeddings generated with Gemini API and cosine similarity to find relevant documents.

Authorization

ApiKeyAuth

AuthorizationBearer <token>

Bearer token for API authentication

In: header

Path Parameters

id*string

Unique ID of the Knowledge Base

Query Parameters

query*string

Search query

Lengthlength <= 1000

limit?integer

Maximum number of results

Default10

Range1 <= value <= 50

similarityThreshold?number

Minimum similarity threshold (0-1)

Default0.7

Range0 <= value <= 1

includeMetadata?boolean

Include metadata in results

Defaulttrue

Response Body

`application/json`

curl -X GET "https://app.docutray.com/api/knowledge-bases/string/search?query=How+to+configure+the+system%3F"

curl -X GET "https://app.docutray.com/api/knowledge-bases/string/search?query=How+to+configure+the+system%3F"

{
  "data": [
    {
      "document": {
        "id": "string",
        "documentId": "string",
        "content": {},
        "metadata": {},
        "createdAt": "2019-08-24T14:15:22Z",
        "updatedAt": "2019-08-24T14:15:22Z"
      },
      "similarity": 0.8542
    }
  ],
  "query": "string",
  "resultsCount": 0
}

Empty