Build an AI Chat Agent in Autom8 AI

In this hands-on guide, you’ll build your first AI-powered chat agent using Autom8 AI. Whether you're new or experienced, this tutorial introduces the building blocks of AI workflows, step-by-step.

🔧 What You’ll Need

  • Autom8 AI Account We recommend using the cloud version (free trial available). Self-hosted users can follow along too.

  • Chat Model Credentials We’ll use OpenAI in this example, but you can also use Gemini, Groq, DeepSeek, Azure, and others.

📚 What You’ll Learn

  • How to use the AI Agent node

  • How to connect a Chat Model

  • How to handle user input and memory

  • How to customize the system prompt

  • How to persist conversations

✅ Step-by-Step Guide

Step 1: Create a New Workflow

  1. Log in to Autom8 AI

  2. From the dashboard, click Create Workflow

  3. Start with a blank canvas

Step 2: Add a Chat Trigger

  1. Click Add First Step or press Tab

  2. Search for Chat Trigger

  3. Add it to the canvas

  4. This node will handle incoming chat messages

Step 3: Add an AI Agent Node

  1. Click the connector on the Chat Trigger node

  2. Search for AI Agent and select it

  3. The agent node will process incoming chat messages with AI

Step 4: Connect a Chat Model

  1. On the AI Agent node, click Add Chat Model

  2. Select OpenAI Chat Model

  3. In the node editor, choose a model (e.g., gpt-4o-mini)

  4. If prompted, add your API key under Credentials

You can get your OpenAI API key from https://platform.openai.com

Step 5: Test the Chat

  1. Click the Chat button on the canvas

  2. Send a message

  3. The AI should respond in real time

  4. The logs on the right show the inputs and outputs

Step 6: Customize the Prompt

  1. Open the AI Agent node

  2. Under Options, add System Message

  3. Change the prompt—for example: “You are a witty assistant who always replies in haiku.”

  4. Save and test again—you’ll see the AI’s tone change!

Step 7: Add Memory for Context

By default, the AI doesn’t remember past messages. Let’s fix that:

  1. On the AI Agent node, click Add Memory

  2. Choose Simple Memory

  3. Set it to store up to 5 interactions

  4. Now, test again—say “My name is Sam”

  5. Then ask “What’s my name?” The AI will remember it

Step 8: Save Your Workflow

Don’t lose your progress!

  • Click Save in the top right

  • Your chat agent is now ready for future improvements

🎉 Congratulations!

You've built your first AI chatbot workflow with:

  • A chat trigger to receive input

  • An AI agent to process and respond

  • A chat model like OpenAI for intelligent replies

  • Custom prompts to shape behavior

  • Memory for context retention

🔍 What’s Next?

  • Add a webhook to deploy your chat externally

  • Connect it to Slack, Telegram, or Discord

  • Enhance it with file uploads or multimodal input

  • Use workflows to turn the AI into a personal assistant, support bot, or creative companion

PreviousIntegrations

Last updated