Ask a question

Submit a question and get a response. It’s mostly a proxy to AI engine request

Parameters


"omni_key" is the private key.
You can generate a personal key in your profile settings on the platform. Instructions on how to do this can be found here> API key

"modelName" is a parameter that determines the language generation model used to process a request or task.

1. gpt-3.5-turbo-1106

2. gpt-4-1106-preview

"proofLevel":
This is a combined value for temperature and top_p. It sets the degree of freedom we allow the model in generating responses—from strictly using project data to completely random. It accepts values from 0 to 1 in increments of 0.1
"temperature" sets the freedom of response generation for the model. Accepts values from 0 to 2 in increments of 0.1.

“prompt”:
"formatting"This custom format can return paragraphs, JSON, or with emoticons.

Example: “subheaders/emojis/bold main keywords/subheaders/bullet points format”
"tone" Sets the tone for formal and informal communication.

Example: “creative/formal/informal/persuasive/collaborative/creative”
"role" for instance”AI adviser/Professor/Sales assistant/William Shakespeare. etc”


"mostLikely" is how to respond if the bot does not know the answer.
"answerLength" Represents the length of the response in characters.

"doubts" You can define custom actions in case the bot is uncertain about the answer. If you include a specific tag (for example, {doubts}) in your query, the chat can then trigger an API request, send an email, or redirect the query to another messaging platform (such as a live agent).”

"followups" is an instruction for the bot to display additional questions that can facilitate the continuation of the conversation.
Example questions:

  1. What other information do you need?
  2. Can I assist you with anything else?
  3. Would you like to explore any related topics?

These messages can be displayed on the UI with a delay of 5 seconds. If the person does not write anything, the bot can automatically insert these messages into the chat window.

A “custom prompt” refers to a personalized input invitation, replacing the standard one. If there’s information to convey, tone and role formatting are disabled, ensuring the priority of the custom prompt. Consequently, if a “custom prompt” is present, it overrides the standard “prompt,” ensuring its exclusive use. Please summarize the original context in up to 100 words, ensuring the entire response is in the original language and formatted using Markdown.
Please take into attention the context: {}
Answer the question: {question}

Example “custom prompt”:

"customPrompt": "{\"question\": \"Write examples.\", \"context\": \"Write examples:\"}",



"language" we support many languages, please indicate which one you use in your project.
Specify the language in code format en/es/ua/de…

"precontext" specifies the number of previous messages in the chat history that are involved in forming the bot’s response.


Request body

/rest/v1/actions/ask/
Example
{
  "openAIApiKey": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VyIjp7ImlkIjoiNGFiNDUyYTMtNGMwYi00MTIyLWI3NDgtYzk4MWRiN2VlYWFlIiwiZGVmYXVsdFJvbGUiOiJ1c2VyIiwiYWxsb3dlZFJvbGVzIjpbInVzZXIiLCJtZSJdfSwiaWF0IjoxNzE0MzgxNzM4LCJleHAiOjE3NDU5MzkzMzgsInN1YiI6IjRhYjQ1MmEzLTRjMGItNDEyMi1iNzQ4LWM5ODFkYjdlZWFhZSJ9.EFa4TEFmXrSdUX7X6rrfbFSy2xP78YEhHSRWUid9hMo",
  "modelName": "gpt-3.5-turbo-1106",
  "proofLevel": 0,
  "temperature": 0,
  "top_p": 0,
  "question": "what is ISTQB?",
  "prompt": {
    "formatting": "subheaders",
    "tone": "creative",
    "role": "Professor",
    "custom": "string",
    "mostLikely": "string",
    "answerLength": 0,
    "history": "string",
    "doubts": "string",
    "followups": "string",
  },
  "language": "en",
  "precontext": "7"
}
Schema
{
  "openAIApiKey": "string",
  "modelName": "string",
  "proofLevel": 0,
  "temperature": 0.0,
  "top_p": 0.0,
  "question": "string",
  "prompt": {
    "formatting": {},
    "tone": {},
    "role": {},
    "customDoubts": "string",
    "followups": {},
    "language": "string",
    "precontext": "string"
  }
}

Responses

Successful operation

Example
{
    "results": [
        {
            "text": "ISTQB stands for the International Software Testing Qualifications Board. It is a global organization that provides certification for software testers. The certification is recognized worldwide and demonstrates a tester's proficiency in software testing principles and practices."
        }
    ]
}
Schema
{
  "results": [
    {
      "text": "string"
    }
  ]
}
Was this article helpful to you? Yes No

How can we help?