Abdullah Yahya, CEO of AdeptiveAI.com, shows us how to build an automated AI chatbot quote system using Voiceflow, Make, and OpenAI.

This chatbot allows businesses to generate quotes in real-time, reducing wait times for potential clients, and improving conversion rates.

When running a business, speed and efficiency are key.

Especially today, customers don’t wait around!

The faster you reply, the better your conversion rates.

A perfect opportunity for AI and automation.

Abdullah deployed this exact chatbot for a large client, reducing their quote turnaround time from HOURS to MINUTES, saving 30 minutes of employee time PER quote request.

Before this chatbot, an employee would manually review a quote request, compare it with previous quotes, then write an email with the new quote and explanation of services. Prospects would wait hours. A terrible customer experience.

After this chatbot, employees simply review a quote draft generated by AI and hit send.

Quick background on Abdullah and how we met:

• He won my Vampire Security Checkpoint game back in August!

• During our 1:1 session, we discussed AI chatbots and automations.

• Since then, he’s mastered building AI, chatbots, and automations for B2B.

• He already has B2B clients in the Netherlands where he resides - even though he just started in August!

Here’s the Youtube walkthrough led by Abdullah:

Table of Contents

• Introduction to Tools

• Step-by-Step Tutorial

  • Step 1: Sign Up for Voiceflow

  • Step 2: Create a New Voiceflow Project

  • Step 3: Design Your Chatbot Flow

  • Step 4: Capture User Input

  • Step 5: Define Intents

  • Step 6: Create Workflows for Each Intent

  • Step 7: Setup Data Capture

  • Step 8: Connect Voiceflow to Make

  • Step 9: Define Parameters in Make

  • Step 10: Create OpenAI Assistant

  • Step 11: Integrate OpenAI Interaction

  • Step 12: Email Integration

  • Step 13: Test, Test, Test!

  • Step 14: Launch and Monitor

• Wrap Up

• Need More Help?

Introduction to Tools

Here’s our toolset:

1. Voiceflow: create AI chatbots without coding. You can create chatbots that automatically engage with customers across different platforms — your website, SMS, or WhatsApp. A single chatbot that responds to customer requests across all channels in real-time, so you can be there for your prospects and customers 24/7.

2. Make.com: A powerful workflow automation tool that connects various apps and services. We’ll connect Voiceflow to Make, which will then call an OpenAI Assistant and send emails via Gmail.

3. OpenAI (ChatGPT): This will analyze a customer’s quote request, compare it to previously generated quotes, and create a similar quote.

Step-by-Step Tutorial

Step 1: Sign Up for Voiceflow

• Visit the Voiceflow website.

• Create a free account to access the dashboard.

Step 2: Create a New Voiceflow Project

• Click “New Agent” at the top right of the dashboard. This starts a new project.

• Name your project “Quote Automation Agent”.

Step 3: Design Your Chatbot Flow

• Open the editor.

• You will see various building blocks on the left side.

• Drag a Talk block into the workspace:

  • Set the welcome message: “Hello! I’m your AI assistant for quote requests. How can I help you today?”

• Drag a Message block with 2 options:

  • Get a Quote or Contact Support, made using Button blocks.

Step 4: Capture User Input

• Set a Capture block to receive the user's input. This allows for flexibility in capturing user queries, in case the user types in a question, instead of selecting a prebuilt button.

• Use triggers to listen for user questions, like “Can I get a quote?”

Step 5: Define Intents

• Within Voiceflow, go to Intents in the left sidebar.

• Create a new intent to detect a quote request.

• Add examples: “I need a quote,” or “How much will it cost?”

• Ensure the confidence bar is filled to 100% by adding multiple examples.

Step 6: Create Workflows for Each Intent

• Create a Get a Quote workflow.

• In this workflow, ask the user relevant questions:

  • "What type of property are you inquiring about?"

  • "How large is the property in square meters?"

Step 7: Setup Data Capture

• For each question, add Entity types to classify the responses.

• For example, categorize “type_of_building” entity as Name. If the user submits a number, Voiceflow will know to ask the question again because the entity should not be a number.

Step 8: Connect Voiceflow to Make

Integration with other tools is what gives AI superpowers.

When a customer submits a request for a quote, our chatbot collects his/her information and sends it to Make, which then sends it to an OpenAI assistant, which generates a precise quote and detailed explanation. Finally, Make sends an email draft to a team member for final review.

The OpenAI assistant crunches numbers using historical quote data to provide an accurate response. This reduces manual work and human error, ensuring customers receive detailed, accurate, and consistent estimates instantly.

• Sign up for Make.com and create a new scenario.

• Add a Custom Webhook to connect data from Voiceflow.

• Grab the webhook URL, which you will copy into the Voiceflow API block.

Step 9: Define Parameters in Make

• In the Make webhook, set up parameters that correspond to the captured data from Voiceflow.

• This includes property type, size, and contact details.

Step 10: Create OpenAI Assistant

If you’re absolutely new to OpenAI assistants, I recommend reading this first.

For our chatbot, Abdullah provided detailed system instructions as well as a JSON file containing historical quote data. When the assistant receives a new quote request, it will analyze the request’s information and compare it to past similar quotes in order to generate the new quote.

Here’s a snippet of the system instructions:

A couple best practices when creating your OpenAI assistant:

• specify the assistant’s role and purpose

• use structured JSON data, instead of unstructured PDF files

• provide a few examples of what ‘good’ looks like

Step 11: Integrate OpenAI Interaction

Now, Make will call our OpenAI Assistant.

• In the Make scenario, add an OpenAI module.

• Use the outputs from the Voiceflow data capture as inputs to OpenAI.

• Create a message that instructs OpenAI to use historical price data and respond with a quote.

  

  Youtube video

• Add a 2nd OpenAI module that instructs ChatGPT to output only valid HTML.

Step 12: Email Integration

• Add an email module in Make.com to send the draft quote to a designated team member to approve before sending to the customer.

• Format the email in HTML to ensure it renders properly formatted in email.

Step 13: Test, Test, Test!

• Go back to Voiceflow and test your chatbot.

• Interact with various prompts to ensure the flow works as intended.

• Confirm the webhook sends data correctly and generates the expected email.

Step 14: Launch and Monitor

• Once satisfied with the functionality, deploy the chatbot on your preferred platforms: website, WhatsApp, Facebook, etc.

• Monitor analytics in Voiceflow and Make to evaluate performance and iterate.

Wrap Up

By integrating Voiceflow, Make, and OpenAI, we’ve created a robust and automated AI-powered quote system. This streamlines the quote generation process, reducing customer wait times, reducing manual work, ultimately increasing conversion rates.

Many businesses remain in the dark about what’s possible with AI + Automation today, thinking it’s too complicated or expensive.

But in only an hour, Abdullah combined Voiceflow, Make, and OpenAI to build this out.

The next steps?

Start experimenting, start small.

Pick a simple use case, build a basic workflow, and iterate from there.

Huge thank you to Abdullah for sharing his knowledge! 🔑 

You can find him on LinkedIn and Skool (first 100 members FREE).

Need More Help?

Whenever you're ready, there are 3 ways I can help you 👇️ 

1. Blotato: overwhelmed creating content? I built Blotato to grow from 0 to 350k+ followers in 6 months. Easily remix any youtube videos, podcasts, blogs, websites, PDFs, etc. into multiple posts. Simplified content creation, at scale.

2. ChatGPT for Beginners: Free course to learn the fundamentals of ChatGPT, starting from ground zero - what is it, how do you install it, etc.

3. AI Automations & Chatbots: Free videos to learn how to build AI chatbots and automations to scale your business and streamline manual work.