POST Agents/AI/Search

• 3 minutes to read
 • 3 minutes to read 
POST /api/v1/Agents/AI/Search

Returns semantic search results from vector database.

Query String Parameters

Parameter Name Type Description
$select string Optional comma separated list of properties to include in the result. Other fields are then nulled out to reduce payload size: "Name,department,category". Default = show all fields. 
POST /api/v1/Agents/AI/Search?$select=name,department,category/id

Request Headers

Parameter Name Description
Authorization Supports 'Basic', 'SoTicket' and 'Bearer' schemes, depending on installation type.
X-XSRF-TOKEN If not using Authorization header, you must provide XSRF value from cookie or hidden input field
Content-Type Content-type of the request body: application/json, text/json, application/xml, text/xml, application/x-www-form-urlencoded, application/json-patch+json, application/merge-patch+json
Accept Content-type(s) you would like the response in: application/json, text/json, application/xml, text/xml, application/json-patch+json, application/merge-patch+json
Accept-Language Convert string references and multi-language values into a specified language (iso2) code.
SO-Language Convert string references and multi-language values into a specified language (iso2) code. Overrides Accept-Language value.
SO-Culture Number, date formatting in a specified culture (iso2 language) code. Partially overrides SO-Language/Accept-Language value. Ignored if no Language set.
SO-TimeZone Specify the timezone code that you would like date/time responses converted to.
SO-AppToken The application token that identifies the partner app. Used when calling Online WebAPI from a server.

Request Body: request

NamespaceName, Query, Type, Top

Property Name Type Description
NamespaceName String
Query String
Type String
Top Integer

Response:array

OK

Response Description
200 OK

Response body: array

Property Name Type Description
Chunk string Markdown text that matched query.
Title string Title or filename of matched document, extracted from the Metadata
Id string Matched document's unique id. Typically a URI.
Type string Content type (web/faq/ticket/...)
Score double Relevancy score (0-1). How close this chunk of the document matched the query.
Metadata object Dictionary of metadata for this result. Keys depend on the source of the document. e.g. 'Title', 'Filename', 'Url', etc.

Sample request

POST /api/v1/Agents/AI/Search
Authorization: Basic dGplMDpUamUw
Accept: application/json; charset=utf-8
Accept-Language: en
Content-Type: application/json; charset=utf-8

{
  "NamespaceName": "Jones-Towne",
  "Query": "sit",
  "Type": "Document",
  "Top": 496
}

Sample response

HTTP/1.1 200 OK
Content-Type: application/json; charset=utf-8

[
  {
    "Chunk": "necessitatibus",
    "Title": "voluptas",
    "Id": "atque",
    "Type": "Document",
    "Score": 26510.505999999998,
    "Metadata": {
      "fieldName": {}
    }
  },
  {
    "Chunk": "necessitatibus",
    "Title": "voluptas",
    "Id": "atque",
    "Type": "Document",
    "Score": 26510.505999999998,
    "Metadata": {
      "fieldName": {}
    }
  }
]