Skip to content
Ishvana Ishvana Docs

Semantic Search

POST
/api/discovery/enhanced-search/semantic-search

Perform advanced semantic search across documents using AI-powered embeddings.

Request Body:

  • query: Search query text
  • top_k: Number of results to return (1-100, default: 10)
  • similarity_threshold: Minimum similarity score (0.0-1.0, default: 0.3)
  • filters: Optional metadata filters
  • model_name: Embedding model reported in the response (default: “multilingual-e5-large-instruct”)

Returns:

  • List of search results with:
    • document_id: Unique document identifier
    • similarity_score: Semantic similarity score (0.0-1.0)
    • metadata: Document metadata (title, tags, etc.)

Raises:

  • 404: No document embeddings found (reindex required)
  • 500: Search operation failed

Request Body required

Section titled “Request Body required ”
SemanticSearchRequest
object
query
required
Query

Search query

string
topK
Topk

Number of results to return

integer
default: 10 >= 1 <= 100
similarityThreshold
Similaritythreshold

Minimum similarity score

number
default: 0.3 <= 1
filters
Any of:
object
key
additional properties
any
modelName
Any of:
string

Responses

Section titled “ Responses ”

200

Section titled “200 ”

Successful Response

Response Semantic Search Api Discovery Enhanced Search Semantic Search Post
Array<object>
SemanticSearchResult

Semantic search result model.

Represents a single document result from semantic search operations including document identification, similarity scoring, and metadata for enhanced semantic search and similarity analysis.

Fields:

  • document_id: Unique identifier of the matched document
  • similarity_score: Semantic similarity score (0.0-1.0) indicating how closely the document matches the search query
  • metadata: Dictionary containing document metadata including:
    • Document title, type, and creation date
    • Project association and tags
    • Content preview and word count

Usage: Used within semantic search endpoints:

  • POST /api/discovery/enhanced-search/semantic-search returns list of this model
  • POST /api/discovery/enhanced-search/similar-documents returns list of this model

JSON Example:

{
  "documentId": "doc_123",
  "similarityScore": 0.87,
  "metadata": {
    "title": "API Documentation",
    "type": "documentation",
    "project": "ishvana"
  }
}
object
documentId
required
Documentid

Unique document identifier

string
similarityScore
required
Similarityscore

Semantic similarity score (0.0-1.0)

number
<= 1
metadata
Metadata

Document metadata

object
key
additional properties
Any of:
string

422

Section titled “422 ”

Validation Error

HTTPValidationError
object
detail
Detail
Array<object>
ValidationError
object
loc
required
Location
Array
msg
required
Message
string
type
required
Error Type
string
input
Input
ctx
Context
object