Documentation
Developer Guide
3rd Party, Docker, and CLI
Using Begin AI Docker Image

Using Begin AI Docker Image

Begin AI Docker Image can be easily used to process live data without the need for any knowledge of python, follow this guide for details on how to use it 🙂

Get the Image

Before pulling the Begin AI Docker Image, make sure you have docker available on your computer, by running the following commands on a terminal:
docker --version
⚠️
If docker is not installed on your machine, install both before proceeding
 
Now, that you have confirmed that docker is available on your computer, let’s pull the image and start a container by running the following commands on a terminal:
docker pull engineeringbeginai/beginai

docker run -d -p {your_exposed_port}:5000 engineeringbeginai/beginai
Once the commands are completed, run the command:
curl http://localhost:{your_exposed_port}/
You should see a "Pong!" message 🙂

Accessing Swagger UI

The Begin AI Docker Image runs a lightweight WEB API built using Fast API, in order to access the Swagger UI, open on your browser: http://localhost:{your_exposed_port}/docs, you should see something similar to:
notion image

Using the WEB API

The WEB API contains two endpoints:
  • / → Simple GET endpoint that only returns "Pong!"
  • /learn → POST endpoint used to perform the learn data process
 

The /learn Endpoint

Header
  • appid → Your Team App ID
  • licensekey → Your Team License Key
 
💡
For details on how to retrieve that information visit
🗂️
Find Your App and License ID
 
Body
This is a sample structure of the expected body:
{
  "object": "string",
  "id": "string",
  "properties": [
    {
      "type": "number",
      "name": "string",
      "value": 0
    }
  ],
  "labels": [
    "string"
  ],
  "interaction": {
    "object": "string",
    "id": "string",
    "action": "string"
  }
}
Mandatory Fields
The object property refers to the name of the object to which you are providing data, the name must follow what was defined in your Schema (for example: book)
The id property represents the id of the object - this could either be a real id in your application or something created just for the integration with Begin AI
The properties field receives a list of fields, for each field you must provide:
  • type → either one of the following options:
    • number
    • text
    • boolean
    • date
    • category
    • id
  • name → the name of the property - must follow the same name as defined in your Schema
  • value → The value to be used (either a number, a boolean or a string)
Optional Fields
The list of labels to be added to the object. This property is optional and should only be used if the labels property is also defined on your Schema
The interaction property is also optional and used to represent an interaction between a user and a user/object, an interaction contains the following properties:
  • object → the name of the object your user interacted with
  • id → the id of the object interacted with
  • action → the name of the action done by the user
💡
Interactions are only available between a user and another object/user
 
For each POST request sent, the WEB API will generate the appropriate embedding and send it over to the Begin AI server but don’t worry the server does not store any data provided