Connecting and configuring AI

To use artificial intelligence on the Gerabot service, you need to take 4 steps.

1. Registration on the OPENAI platform and creation of a token to use ChatGPT in the chatbot.
2. Adding a token in the AI ​​settings on the Gerabot platform to synchronize the bot with ChatGPT.
3. Creating rules for artificial intelligence (AI training).
4. Implementation of AI in the chatbot structure.

Registration and receiving a token.

To receive a token, you need to create an account on the site - https://platform.openai.com/, create a new project, add a payment option. After that, in the "Your profile" tab, create an api token for the project.

Adding a token to Gerabot.

To synchronize Gerabot and ChatGPT, the received token must be added to the system in the section “Advanced Features”->”Artificial Intelligence. ChatGpt”.

Creating rules for AI work.

In order for the AI ​​to provide answers to users according to the bot's needs and topics, it is necessary to create rules (instructions) for the work of the AI. The bot can use different rules of AI operation, depending on the exact moment of the bot's operation when it is activated to interact with the user. The "Create rule" button is located in the "Response generation" tab.

Creating a rule.

List and description of settings for creating a rule.

Name. 

A system field that is not visible to the bot user.

Instructions for the bot (Prompt).

The key setting in which you need to provide instructions to the AI ​​on which topic it should respond to users. The more precise and detailed the instruction, the better the answers from the AI ​​will be. Also, in this setting you can write, for example, which topics the AI ​​should avoid when talking to the user. 

Example 1: The company provides logistics services. Answer general questions on this topic, provide information and contact details of company managers for more detailed clarification.

Example 2: Topic of answers - seo site promotion. Provide the user with general information on the topic, if possible, support the information with numbers. Avoid the topic of contextual advertising.

Message after launch. 

The first message that the user will see after switching to AI communication.

Model.

The main difference between the models is the amount of response, data processing speed and data relevance. 

The most popular models for working with non-English content are gpt-4o-mini and gpt-4o.

Maximum number of tokens for responses.

Maximum amount of AI response to a user's request (on average 4 tokens = 3 words).

Limit usage per chat.

The maximum number of tokens that can be used by one chat per day.

Keyboard to continue the bot.

To exit the user from the mode of operation with artificial intelligence, the bot may include a keyboard with a button to continue working with the bot.

Scenario after the limit

The script that will be displayed to the user when the limit of used tokens is reached.

Add action to keyboard

In order to  in order for the script keyboard buttons to be able to run the artificial intelligence module, this option must be enabled.

Adding the launch of the AI ​​module to the bot.

When the AI ​​rules are created, you need to choose a script and a button that will start the AI ​​for the user. To do this, you need to go to one of the created scenarios (or create a new one) in the Scenarios section, and in the keyboard tab, create a button and add the corresponding action when pressed to it - launch AI with the corresponding rule.