Conversational AI API Reference

Sherpa.ai Conversational AI is a REST API that provides developers with a framework and toolbox to formulate any request for Sherpa.ai, by using standard syntax technologies like HTTP Rest Services with a predefined set of URLs. These URLs represent various resources, any information or content accessed at that location, which can be returned as JSON, HTML, audio (speech) files or text.

API Endpoints

Sherpa.ai Conversational AI is a set of HTTP endpoints that help your app integrate with Sherpa.ai. These endpoints accept arguments as JSON in the request body and return results as JSON in the response body.

All Sherpa.ai Conversational AI API endpoints will be served on https://api.sherpa.ai/v2

Dialogue endpoints will be user-tailored, making use of the RequestInfo header parametrization, prioritizing location over other results.

Header Parameters

All the API calls have common parameters in the header.

The specific header parameters will be defined on each API call.

Sherpa-Request-Info

Specific Sherpa.ai header where parameters are grouped in the form of a JSON object.

Besides the specific information to be sent to each endpoint, Sherpa-Request-Info will gather the request context as a request header field.
The fields in RequestInfo are optional. However, each additional piece of data helps Sherpa.ai fetch a more tailored response, so including as many fields as possible is recommended.

For some location-related requests (weather, restaurants, etc.), the latitude and longitude are required. If they are not included, no results will be returned.

For example, by setting the lat and lon fields to specify user location, Sherpa.ai can use that information to respond to queries such as "Where am I?" or "Recommend a restaurant near here."

Example

In order to make a text-based dialogue query regarding the weather, latitude and longitude parameters are mandatory in the RequestInformation data header.

Request:

curl -X POST "https://api.sherpa.ai/v2/query/text" \
-H  "Authorization: Basic XXXX-SHERPA-TOKEN-XXXX" \
-H "Sherpa-Request-Info:{\"lat\":37.468319,\"lon\":-122.143936}" \
-H  "Accept-Language: en-US" \
-H  "Time-Zone: Europe/Madrid"
-H  "accept: application/json" \
-H  "content-type: application/json" \
-d "{\"q\":\"what's the weather like?\"}"

Be aware that if you are using a tool like POSTMAN or PAW, you don't need to escape the Sherpa-Request-Info JSON, as the tool does it for you.

Response:

{
  "context": {"ref": 342673096},
  "nextActions": [{"data": ""}],
  "responses": {
    "query": "what's the weather like?",
    "text": "Right now, in East Palo Alto, it is 51 degrees and cloudy. The temperature will rise in the coming hours."
  }
}

Parameters

Each endpoint method will have a parameter table with explanatory information, along with an example. In this table, information will be provided for each field, like 'source', 'mandatory', 'type', and 'description'.

• Field
  Parameter field

• Source
  Where the parameter has to be placed

  • Query
    Parameter is part of the URL

  • Body
    The parameter is set inside the body payloads part of a POST/PUT type method request. These parameters will be in JSON format and will require the corresponding header to be set:

    -H  "content-type: application/json"

• Mandatory
  Whether parameter is mandatory or not
• Type
  Value type [String, Integer, etc.]
• Description
  Additional information