How to Create Your First AI Agent (for Non-Coders)

A comple guide to create an AI Agent for beginners.

For many people, building their first AI Agent can seem like a challenge. The first thing that often comes to mind is that it involves a lot of coding. While that’s true, there's also the option to create an agent using low-code platforms. These are usually more intuitive and provide a faster way to understand how an agent works, without requiring a heavy coding background.

In this guide, we’ll build our first AI Agent using the best tool: n8n

In this beginner tutorial, we’ll create an AI Agent that can send us an email with information about five cryptocurrencies. Later on, we'll see how we can also chat with the agent and ask it to retrieve information it has generated.

Let’s get started!

Table of content:

- Step 1: Set Up Your Workspace
- Step 2: Add the AI Agent and Configure Its Brain
- Step 3: Connect External Tools for Automation
- Step 4: Teach Your Agent and Launch
- Step 5: Handle Common Errors
- Step 6: Run and Interact with Your AI Agent

Here’s the link to the n8n platform. The first thing to do is create an account. Once you're in, we'll start by creating a new project.

Step 1: Set Up Your Workspace

We’ll customize things later based on the type of development we plan to do.

To start, go to the Workflows tab and select Start from scratch.

Click on Add first step. A list of available triggers will appear. Choose “On a schedule”

This trigger lets you set how often the workflow should run, based on your preferred schedule.

Step 2: Add the AI Agent and Configure Its Brain

Next, click the plus icon (+) to open the options menu. For this project, choose AI, and then select AI Agent.

Once the AI Agent node is added, you’ll see a few different sections. On the left side is the input to the node. In some workflows, this could be the output from a previous step, but in our case, it’s coming directly from the trigger we just set up. On the right side is the output, which will be passed to the next node once this one finishes. In the center, you’ll find the parameters and settings where you can configure the behavior of the agent.

Back on the canvas, you’ll notice the trigger is now connected to the AI Agent node. These connections between nodes can be adjusted at any time. You can remove them or create new ones depending on your workflow. The AI Agent node lets you set up the LLM (Large Language Model) you want to use. n8n provides three key components to work with here: Chat Model, Memory, and Tool.

Click the plus icon (+) next to the Chat Model section. A panel will appear on the right side showing the list of available LLMs. From there, you can choose the one that best fits your project.

There are a variety of models to choose from, depending on the goals of your workflow. For this project, I’ll go with OpenAI, which provides free usage credits if you’re new to n8n.

Using OpenAI also gives you access to several models. For this example, we’ll use GPT-4 mini.

At this point, the core of your project is in place.

Next, we’ll configure memory for the workflow. Click on the Memory node. On the right side, you’ll see a list of options. Choose Simple Memory, which works well for the kind of setup we’re building.

In simple terms, the Memory setting tells the LLM how many previous messages it should remember during the workflow.