📃Adding Sources

Sources are the data with whom you will train your bots. The amount of data you can use (in terms of words) depends on your account Plan.

Data Types

For now, you can use data's coming from website (URLs) or documents (pdf, doc, docx...).

For PDFs, even scanned files can be used thanks to a solution called tesseract.

Tesseract OCR is an open-source optical character recognition engine that is widely used for converting images of text into editable and searchable documents

Web sources (links)

The endpoint to add Links as sources is : https://www.owlbot.ai/api/weblinks/learn

This endpoint only accept POST Request with the good authentication token as explained in the dedidcated section.

Here is the parameters you have to provided :

Property	Type	Description
links	array	Required. like `["link1", "link2"]`
chatbotId	uuid	Required. It's defined the chatbot UUID of the chatbot you want to feed with thoses links.

Remember that the method used to extract data from the links you provide is called scrapping.

As per the general rules of internet etiquette and legal standard, you are only allowed to scrape data from websites that you own or for which you have explicit written permission from the owner. The unauthorized scraping of data from a website that you do not own is considered a violation of copyright laws and the website's Terms of Service.

Web scraping may seem harmless, but it can have serious consequences. Unauthorized web scraping can lead to legal repercussions, including potential lawsuits for copyright infringement and privacy violations. Moreover, it can burden the web servers, leading to performance issues and even outages, which negatively impact all users.

Adding File / Document

The endpoint to add Links as sources is : https://www.owlbot.ai/api/upload

We do support POSTing binary files directly to our storage space. You have to use formData to do so.

Property	Type	Description
file	binary	Required. As to be 20 MegaBytes maximum.
name	string	Required. the filename that identify this file.
type	string	Required. FOr now the supported file type are listed below
chatbot	uuid	Required. It's the chatbot UUID of chatbot you want to feed with thoses documents.
storage	string	Required. as to be "private" because your uploaded files are not available for other users

Response

Response is a HTTP 200 with a JSON source object :

{ "uuid" : "dc94a8c3-7cab-4452-be1e-f2c029132216" }

Where uuid is the unique identifier of the uploaded file.

Some examples :

In Javascript / Fetch

var myHeaders = new Headers();
myHeaders.append("Authorization", "abcdefghf");

var requestOptions = {
    method : "POST",
    headers : myHeaders,
    redirect : 'follow'
};

const fileForm = new FormData();
 fileForm .append("file", _file);
 fileForm .append("type", _file.type);
 fileForm .append("name", _file.name);
 fileForm .append("storage", "private");
   
 fetch("https://www.owlbot.ai/api/endpoint", fileForm ,requestOptions)
  .then(response => response.text())
  .then(result => console.log(result)
  .catch(error => console.log('error', error));

Supported File Type

Supported file	MIME Type
Word File	"application/vnd.openxmlformats-officedocument.wordprocessingml.document", "application/wps-office.docx", "application/msword"
PDF File	"application/pdf"