Help

Artificial intelligence

AI agent, training, and functions.

16 articles

articles

How to configure an AI agent in the chatbot

Link an AI agent to the chatbot flow, collect user attributes, and set function exit conditions.

How to configure the conversation tab

The conversation tab controls which media types the agent processes and how it formats responses.

How to configure the general tab of the AI agent

The general tab sets the model, credentials, and behavior parameters for the agent. The OpenAI token is required. Pinecone is only needed if you plan to.

How to create an AI agent on the platform

Create the agent under Modules, give it a descriptive name, and link the authorized companies.

How to create and access your Pinecone token

The Pinecone token connects the AI agent to the long-term memory database. Save the token immediately after creation: it is shown only once.

How to create and configure a function for an AI agent

Functions define specific actions the agent can execute, such as querying a database, sending notifications, or collecting user data.

How to set AI agent instructions

The Instructions tab defines who the agent is, how it speaks, what it does and the goal. That drives behavior in chats.

How to set function return in the chatbot

After the agent runs a function, the block defines what the flow does next: return to the agent or stop there.

How to use the Training tab

The Training tab feeds the AI agent with links, files, and question-and-answer pairs.

Module: Function return

The function return block closes the execution of a function flow and defines what happens next.

New automation menu

The Automation menu consolidates all automation settings in one place on the platform sidebar.

New agent settings screen

The General settings tab controls the base behavior of the agent. Access it by clicking Configure on the agent card.

Tab: Extra context info

Use this tab to enter additional information about the current contact. This information enriches the agent's responses during the conversation without.

Tab: Input instructions

Use this tab to define specific instructions for the agent based on the flow block the user is coming from. The agent adapts its initial behavior.

Tab: Max wait time and limits

Set time and resource limits for the agent. These parameters prevent loops, control token consumption, and avoid excessive requests.

Training guide: web pages and subpages

In the agent's Training tab, the Page scraping limit field defines how many pages on your site the agent can access to extract information.