Building Your First Chatflow

[Intro]

All right, welcome back. This is where things get exciting. Oh yes, because we are diving into something handson. Building your first chat flow. But before we jump in, if you don't have your custom account yet, head over to custom.io and get started. Seriously, it takes only a few minutes and you will be ready to follow along with this lesson. Now, let's talk chat flows. Chat flots are the key to creating smooth, structured conversations, automating repetitive tasks, and ensuring that customers are seamlessly routed to the right person or AI for help. Think of chat flows as the GPS for your customers interactions. They guide the journey, adapting to customer needs while keeping things efficient and on brand.

Perfect. Let's get started by creating two chat flows. The first is a basic flow where Aura handles all incoming conversations and the second is a more advanced flow that uses customer intents to guide responses and assign conversation to the right team. So what are chat flows? Before we build, let's

[What Is a Chatflow?]

clarify what chat flows do. Chat flows are customizable chat experiences that allow you to automate repetitive task, browse conversation based on triggers like a customer's questions or intent, and escalate smoothly to a human agent when the AI can fully resolve the issue. With the Orura's AI integrations, chatbots can recognize customer intents, respond intelligently, and end off conversations when needed. They're smarter, faster, and more flexible than traditional chats. Let's build your

[Build your first AI Chatbot]

first chat flow. So we will start with a simple flow where aura is the first point of contact for every common chat. Now go to chat flows in the custom dashboard. This icon over here and let's click on the plus. You should already have it, but for the sake of this tutorial, we're going to create a new one from scratch. So you will learn the basic. So when you click on this, you will have plenty of templates to choose from. And basically, we're going to build this one. Let Aura AI answer first. But we're going to create it from scratch. So you will learn the steps that takes into building a chat flow.

So when you start from scratch, you will need to decide which is the starting point of this chat flow. What should be triggering this chat flow, right? Uh so the first is going to be AI recognized intent. This is something that we're going to define later in our second chat flow. For this example, we're going to use the trigger when a new conversation started. Now you can click on next. Here we can also narrow down the audience that is going to enter in this chat flow. What that means that basically you can narrow down a specific customer segment. For example, all your basic customers or all your users or your leads or whatever you you choose the property or the kind of events they've been triggering on your platform.

You can decide which people should enter this chat flow. Now there is a dedicated lessons for that. So let's for now keep it to open to everyone. Okay. So let's not add any filter here and save and close. Now what shall we do? Whenever a new conversation starts, you can build almost anything over here. You can send a message. You can send a choice. But this is um what traditional chatbots do. So please don't do that. You can start with the AI assistant. When you click the assistant, that will be it. And as you can see, Aura is going to gather the resources from the app center article or the candidate responses.

Now, we got plenty of intents over here to play with, but this is not the case. We will see them later. For the sake of this tutorial, the only thing you have to do is to close this panel over here. Now we need to decide what to do with the conversation that Aura is going to solve and what to do with this conversations that Aura needs to escalate to someone because she couldn't get there. Now we need to decide what to do with the conversation that Aura assault and now you can click on this. Now most often you can just close the conversation and that will be enough because in this way also you will have the seesant rating going out and you can track performance something that you will learn how to do in one of the next lessons.

Now just click on logic and close conversation. That's great. That's the only thing you have to do for now. We are going to just assign this to your based customer support team. That would be it. You can rename this transfer with AI first to aura and you can publish this.

[Try your AI Chatbot]

Now you can immediately preview this which will be opening a fake website. You can try your chatbot. You can start a conversation and see how that will perform. Once your chat flow is ready, test it to ensure aura responds correctly and escalation work as expected. This flow is perfect for automating initial customer interactions while ensuring smooth handoff when needed. Now, let's create a chat flow

[Build AI recognized intent Chatflow for Sales]

that routes conversations based on specific intents. For instance, you can set up a flow to recognize pricing inquiries and sales demo requests and connect those customers directly with your sales team. Here's how to do it. Let's create a new chat flow. So, again, let's click on plus. Let's create one on our own over here. Now we need to choose the triggering events over here and this time we're going to use AI recognized intent. Now we click on next and we want to identify pricing inquiries. Okay. So this is just for uh internal purposes. So we're going to type down over here pricing. And here this is very important because it's a description for AI on how to recognize this intent.

Okay. So you should tell the AI what to expect. we can u define it as um a user is trying to uh understand the pricing option. That would do it. So here there are two things that you can do to better refine what aura is going to match or not match as this agent. For example, uh questions like um what's your pricing for seat can be a matching agent. And remember these are not checked against like as they are. So of course aura can understand variations of this but this is only to share a bit of more examples when requests get more complex or differs from the very main topic.

Okay, this is very simple. So it's going to match quite easily. But let's say I don't want people to match this intent when they're asking for upgrade. Okay, so this is only for pricing. So in this case, we're going to add over here a non-matching intent example where customer ask how do I upgrade? Okay, cool. So what to do with this pricing intent? As for right now, we will make it super simple. Okay, so every new conversations coming from the chat are going to be assigned to a human being or the sales team. Let's assign this to the sales team. Okay, over here and close it.

So the only thing that this chat flow is going to do is whenever uh Aura is recognizing a pricing question is going to bug specifically the sales team. Okay, we can do way way much more with sentence recognition AI mission, but this is something that we'll cover down the road in this master class. Okay, I'll promise you you will learn and you will be astonished by what Ora can do on your own terms. Okay. So, for the sake of this tutorial, let's save this as pricing queries intent. Okay. And let's publish it. Okay. Cool. So, we're going to test it out again on the very same page that you were being using.

Okay. So, once you deploy it, once you publish it, Ora will recognize this in at every step. So, it doesn't matter if the user started a conversation with a query and now he wants to understand the pricing. authorizing to add the to add 1,000 invoices to my account. Now, what it should happen that will recognize the intent and you will notice that now she stopped typing for a sec because she identified the intent and escalated the conversation. Now you can go and verify this. You can go under your inboxes and now if you can see over here there is a new conversation and if you check the history of what happened.

So the very first message triggered the full scale AI support autoonation chat flow that we've been having. But now this is the intent that triggered the pricing query intense that we created before. This is mind-blowing because right now you can provide a series of answers to the questions until the AI eats a certain intent. Okay? Or it's something different. This is so powerful and you will see later on in this course how to exploit this intent plus missions plus integration to achieve almost 80% of customer support automation. So this setup ensures customers are directed to the right resource quickly improving efficiency and satisfaction. Chat flows are super customizable thanks to the

[Chatflow Triggering events and actions]

triggering events and actions. Here are the two common triggers you can use. First, the new conversation activates when a customer starts a new chart and second intent recognition, which basically triggers when Aura detects a specific customer needs like the pricing that we just saw. actions instead guide the conversation and ensure smooth end off like the assign to a specific team like send a message or you can collect information or you can escalate to a human or you will see you can integrate with external services to bug your team on Slack or send an email or send the data to another service that you want to connect with. Before you wrap up, keep

[AI Chatbot Tips & Tricks]

these tips in mind. Keep it simple. Overly complex flows can confuse both you and your customers. Start simple and build as needed. Second, review and update regularly. Customer needs evolve, so your chat flows should too. Use data and feedback to improve them over time. Plan for multiple languages. If you serve a global audience, set up separate flows for each language. This ensures every customers get responses in their preferred languages. By the way, if you go AI first, you don't need to create different chat flows to provide support in different languages because Aura can talk any language and she adapts automatically to your customer language by default. Use filters for targeting. Customary filters allow you to segment your audience and create tailored chat flow for specific groups like new customers or returning subscribers. Here's what we covered. Chat flows help

[Wrapping Up & Next Steps]

you create structured automated conversations. We built two examples. A flow for handling all conversation with Aura and a flow for identifying specific intents. With triggers like new conversation and intent recognition and actions like assigned to the team by using chat flows, you're not just automating support. We are creating an efficient engaging experience that keep customers happy. What's next? In the next lesson, we will explore assistant versus intent and dive into the differences between AIdriven responses and intentbased flows. If you haven't already, head to custo.io and start your free trial today so you follow along with all these tools. See you there. Ciao.