Conversation Flow

Conversation Flow outlines how the conversation progresses from one point to another, typically organized in a tree-like structure with nodes representing different states or steps in the conversation.

​

Main elements in conversation flow

​

1. Nodes

Nodes represent individual states or steps within the conversation. Each node typically corresponds to a specific action, prompt or response.

​

2. Connections

Connections specify the pathways through which the conversation moves forward, allowing for branching, looping or other forms of interaction.

You have already created an Bot Stack and added your data to the knowledge base, let’s build a simple assistant using the nodes and connections.

On the AI Assistant Designer you will have the Node Stack on the left side, here you will find all the nodes that you can use in the designer to build your AI Assistant. Each node has it’s own purpose and you can learn more about them by clicking here

For now we will only focus on four nodes

1. Start Node
2. LLM Node
3. Response AI Node
4. Listen Node

​

1. Start Node

The start node represents the beginning of the conversation. It is triggered by a mention or from the initial message in a DM. The input provided that initially started the conversation is immediately passed onto all the start nodes outputs.

​

2. LLM Node

The llm node utilizes the Large Language Model of your choice, such as OpenAI GPT, Anthropic Claude to handle conversational dialogue.

​

3. Response Node

The Response AI node takes input from the nodes in the workflow and sends it to the user on chat

​

4. Listen Node

The listen node waits for user input within the chat. Once a message is received, it is immediately passed onto all the Listen nodes outputs.

Here we have added the four nodes we mentioned above, and connected them in a sequence

The conversation flow begins from the Start Node, then transitions to the LLM Node, followed by the Response AI Node and finally, it moves to the Listen Node. By connecting the Listen Node to the LLM Node, it forms a loop, ensuring that the assistant won’t end the conversation but rather continue the flow indefinitely.

You can copy the settings on LLM Node from the above image. If you want to know more about how the LLM node works, click here

Now you can proceed to the next chapter, which covers how you can test the assistant that you just built.