Build a 24/7 AI Employee for Your Business: A No-Code Video Tutorial Guide

Posted on by
Contents show
Build a 24/7 AI Employee for Your Business: A No-Code Video Tutorial Guide
Build a 24/7 AI Employee for Your Business: A No-Code Video Tutorial Guide

Build a 24/7 AI Employee for Your Business: A No-Code Video Tutorial Guide

Imagine having an employee who works around the clock, never takes a break, and responds to customer inquiries instantly. This employee understands your business, answers questions accurately, and maintains a professional, friendly tone every single time. This isn't a scene from a science fiction movie; it's a practical reality you can build for your business today, without writing a single line of code.

Many businesses, especially small to medium-sized ones, find themselves inundated with repetitive customer emails. Questions about store hours, product details, pricing, and shipping policies can consume hours of valuable time. For a small operation like a local bakery or a boutique online store, handling this volume manually can be a significant drain on resources. This guide will walk you through creating a dedicated AI agent that can manage these tasks for you, freeing you up to focus on growing your business.

We will use a powerful visual automation platform called Make to construct this AI assistant. You will learn how to give your agent a ‘brain' by connecting it to an advanced language model, provide it with ‘tools' to access your business information, and set up a complete workflow that automatically handles customer questions from start to finish.

Introducing Make: Your Visual Automation Workshop

Make is a platform that lets you connect different apps and automate workflows using a simple drag-and-drop interface. Think of it as building with digital LEGO blocks. Each block represents an action in an app, and you connect them to create a sequence, or ‘scenario,' that runs automatically.

The platform has recently introduced a feature called AI Agents, which takes this automation to a new level. Instead of just following a rigid set of instructions, these agents can think, reason, and decide which actions to take based on the information you give them. This allows for far more dynamic and intelligent automation. To follow along with this tutorial and build your own agent, you'll need a Make account.

[GET STARTED WITH YOUR FREE MAKE ACCOUNT]

[GET STARTED WITH YOUR FREE MAKE ACCOUNT]

Phase One: Creating the Brain of Your Operation

The first step is to build the core of our AI agent. This involves defining its identity, connecting it to an AI model that gives it reasoning abilities, and providing it with its primary set of instructions.

Setting Up Your First AI Agent

Once you have signed into your Make account, you will land on the main dashboard. On the left-hand navigation panel, you will find an option labeled “AI Agents.” This is your command center for creating and managing your digital workforce.

Click on “AI Agents” to enter the creation space. You can start a new agent by clicking the “Create agent” button, either in the center of the screen or in the top-right corner. This opens a configuration window where we will begin to shape our new employee.

Connecting to an AI Model with Groq

An AI agent needs a Large Language Model (LLM) to function. The LLM is the engine that processes language, understands context, and generates human-like responses. Make supports connections to numerous AI model providers, including well-known names like OpenAI and Anthropic.

For this guide, we will use a provider called Groq. Groq is notable for its incredible speed and for offering free access to powerful open-source models, including Meta's new Llama 3. This combination of speed and capability makes it an excellent choice for a responsive customer service agent.

To connect Make to Groq, you need an API key. You can get one for free by following these steps:

  1. Navigate to the Groq website (groq.com).
  2. In the top-right corner, find the “Dev Console” button and click it.
  3. Sign up for a free account using your email, GitHub, or Google account.
  4. Once signed in, click the “Create API Key” button.
  5. Your new API key will be displayed. Copy this key immediately, as it will not be shown again.

Back in the Make “Create agent” window, you'll see a prompt to create a connection. Click the “Create a connection” button. From the “Connection type” dropdown, select Groq. Give your connection a name, like “My Groq Connection,” and paste your copied API key into the “API Key” field. Click “Save” to establish the connection.

Defining Your Agent's Purpose: The System Prompt

With the connection active, we can now configure the agent itself. First, give your agent a name that reflects its role, such as “Customer Service Agent.” Next, you need to select the specific AI model it will use. Groq provides several versions of Llama 3. The “llama-3.3-70b-versatile” model is a great all-around choice, balancing speed with a deep understanding of complex requests.

The most critical part of this phase is writing the “System prompt.” This is the agent's job description and core set of instructions. It tells the agent who it is, what its purpose is, and how it should behave. The more detailed and clear you are here, the better your agent will perform.

A good system prompt defines the agent's persona, its primary function, and any constraints or rules it must follow.

Here is an example of a detailed system prompt for our Kevin Cookie Company agent:

  • “You are a friendly and helpful customer support agent for the Kevin Cookie Company.”
  • “Respond to customer questions about things like cookie flavors, store hours, shipping, and pricing.”
  • “Keep your tone warm, polite, and professional — like you're talking to a real customer.”
  • “Answer clearly and concisely.”
  • “If a question can't be answered, politely let the customer know and suggest contacting support.”

This prompt establishes a clear personality and provides explicit instructions on how to handle various situations. Once you've written your system prompt, click “Save” in the bottom-right corner. You have now officially created your AI agent's brain.

Phase Two: Equipping Your Agent with Tools for Success

An AI agent with a brain is a good start, but it needs tools to perform useful tasks. It cannot answer questions about your business if it doesn't have access to your business's information. In Make, these tools are built as ‘Scenarios' that the agent can call upon when needed.

We will create two essential tools for our customer service agent: one to retrieve information from a company FAQ and another to send emails to customers.

Tool #1: Building a Knowledge Base with Google Docs

A frequently asked questions (FAQ) document is a perfect source of knowledge for a customer service agent. Let's assume the Kevin Cookie Company has a Google Doc that lists common questions and answers about its products and services. We need to build a tool that allows our agent to read this document.

  1. Navigate to the “Scenarios” section from the left-hand menu.
  2. Click “Create a new scenario” in the top-right corner.
  3. In the center of the visual designer, click the large plus icon to add your first module.
  4. A list of apps will appear. Find and select “Google Docs.”
  5. From the list of actions, choose “Get Content of a Document.” This module does exactly what its name suggests.
  6. Connect your Google account if you haven't already. Then, select the specific Google Doc that contains your FAQ.
  7. Next, we need to define what information this scenario will provide back to the agent. Add another module by clicking the plus icon next to the Google Docs module.
  8. Search for and select the “Scenarios” module. Choose the “Return output” action. This module is used to send data out of a scenario, making it usable as a tool.
  9. Click “Add scenario outputs.” Here, you'll define the structure of the data being returned. Click “Add item” and give your output a name, like “KCC_FAQ,” and a clear description, such as “Kevin Cookie Company FAQ.” The type should be “Text.”
  10. Click the “Scenarios” module again. In the field for your newly created output (“KCC_FAQ”), map the data from the Google Docs module. Select “Text Content” from the list of available data points. This tells the scenario to return the full text of the FAQ document.
  11. Finally, this tool should only run when the agent needs it. At the bottom of the screen, click the scheduling icon and set the scenario to run “On demand.”
  12. Give your scenario a descriptive name in the top-left corner, like “Kevin Cookie Company FAQ,” and save it.

You now have a tool that can instantly provide your company's FAQ to your AI agent.

[BUILD YOUR OWN AI AGENT NOW]

Tool #2: Giving Your Agent a Voice with an Email Scenario

After finding an answer, the agent needs a way to deliver it to the customer. We'll build a second tool that allows the agent to send an email. This tool will require three pieces of information to function: the customer's email address, the email subject, and the message content.

  1. Create another new scenario.
  2. Before adding any modules, we need to define the inputs this scenario expects. At the bottom, click the “Scenario inputs and outputs” icon.
  3. Under “Scenario Inputs,” click “Add item” three times to create three input fields.
  4. Name them appropriately: “Customer_Email_Address,” “Email_Subject,” and “Email_Content.” The type for all three should be “Text.” Save the inputs.
  5. Now, add a module to the scenario. Search for and select your preferred email application (for example, “Email,” “Gmail,” or “Microsoft 365 Email”).
  6. Choose the action “Send an Email.”
  7. Connect your email account if needed.
  8. Now, map the scenario inputs to the corresponding fields in the email module. Click in the “To” field, and from the variables panel, select the “Customer_Email_Address” input. Do the same for the “Subject” and “Content” fields, mapping them to the “Email_Subject” and “Email_Content” inputs.
  9. Set this scenario to run “On demand,” give it a clear name like “Send email replies to customers,” and save it.

Your agent is now equipped with a knowledge base and the ability to communicate.

Linking Tools to Your Agent

The final step in this phase is to tell your AI agent that these new tools are available for it to use.

  1. Return to the “AI Agents” section and open your “Customer Service Agent” for configuration.
  2. Scroll down to the “System tools” section and click “Add.”
  3. You will see the two scenarios we just created. Check the box next to each one.
  4. For each tool, you must add a description. This is extremely important. The description tells the agent what the tool does and when to use it.
    • For the FAQ tool, a good description would be: “Retrieves the Kevin Cookie Company's FAQ from a Google Doc. Use this to answer common customer questions.”
    • For the email tool, you could write: “Sends an email reply to the customer using the provided email address, subject, and message content.”
  5. Add both tools and save your agent configuration. Your AI agent is now fully equipped and ready for action.

Phase Three: Putting Your AI Agent to Work

With the agent and its tools prepared, it's time to build the main workflow that will connect a customer's question to the agent's automated response. This workflow will be triggered whenever a customer submits a question through an online form.

Designing the Master Workflow

For this example, we'll use a form created with Tally, a simple and free form builder. The form will collect the customer's name, email address, and their question. Our master scenario in Make will start with a trigger that “watches” for new Tally form submissions.

  1. Create a third and final new scenario.
  2. Click the plus icon to add the trigger module.
  3. Search for and select “Tally.”
  4. Choose the “Watch New Responses” trigger.
  5. You'll need to create a webhook to connect Tally to Make. Make provides a unique webhook address. Copy it. In your Tally form's settings, go to the integrations tab and add a new webhook, pasting the address from Make.
  6. Once connected, Make will listen for new submissions to that form.

Running the AI Agent and Delivering the Response

The trigger is set. Now we add the action: running our AI agent.

  1. Add a new module after the Tally trigger.
  2. Search for and select “Make AI Agents.”
  3. Choose the action “Run an agent.”
  4. From the dropdown menu, select your “Customer Service Agent.”
  5. Now we need to provide the agent with a prompt for this specific task. In the “Messages” section, click “Add item.” This is where you will construct the instructions for this particular run.
  6. Craft a clear message using the data from the Tally form. The panel on the right will show all the data received from the Tally trigger, including the fields for “Your Name,” “Email Address,” and “Question.”

Your message could look like this:

  • Customer Name: [map the “Your Name” field from Tally here]
  • Customer Question: [map the “Question” field from Tally here]
  • Customer Email Address: [map the “Email Address” field from Tally here]
  • “Can you respond to this question via email?”

This dynamic prompt gives the agent all the context it needs: who the customer is, what they asked, and their contact information. Based on this prompt and its system instructions, the agent will decide to use its tools. It will first use the FAQ tool to find the answer and then use the email tool to send the composed response.

After configuring the agent module, turn the scenario on so it runs immediately whenever new data arrives from Tally. Give the scenario a name, like “Integration Tally, Make AI Agents,” and save it.

[SIGN UP FOR MAKE FOR FREE AND BUILD YOUR AI AGENT TODAY]

The Final Test: Witnessing Automation in Action

The entire system is now built. To test it, simply go to your Tally form and submit a question as a customer would. For example, a customer named “Cookie Monster” might ask, “What are your store hours on Saturday?”

Within moments of submitting the form, the Make scenario will trigger. The AI agent will receive the prompt, access the FAQ document to find the store hours for Saturday, compose a polite and helpful email, and use the email tool to send it to the customer's address. The resulting email will be personalized and accurate, providing a seamless customer experience.

Beyond Customer Service: The Versatility of Make AI Agents

This customer service bot is just one application of Make's AI Agents. The true strength of the platform lies in its flexibility. Because Make integrates with thousands of applications, you can build agents for nearly any business process.

Imagine an AI agent that:

  • Monitors your social media for mentions and drafts replies for your approval.
  • Takes new leads from a CRM, researches them online, and adds enrichment data back to the CRM record.
  • Analyzes sales data from a spreadsheet each week and emails a summary of key trends to your team.
  • Manages your calendar by processing meeting requests from emails and scheduling events based on your availability.

By combining different triggers, tools, and a well-crafted system prompt, you can design specialized AI agents for marketing, operations, sales, and more. Each agent can be equipped with a unique set of tools tailored to its specific function, creating a truly customized and powerful automation ecosystem for your business.

You have the framework to build a sophisticated, autonomous assistant that can handle complex, multi-step tasks. You can add more tools, refine your prompts, and connect more applications to expand your agent's capabilities over time.

[SIGN UP FOR MAKE FOR FREE AND BUILD YOUR AI AGENT TODAY]

More Articles for you:

FTC Affiliate Disclaimer: This article contains affiliate links. I may earn a commission from qualifying purchases made through these links.

[SIGN UP FOR MAKE FOR FREE AND BUILD YOUR AI AGENT TODAY]
Subscription Form