All right, so now we're going to be going through each and every individual node that currently exists. I'm just going to walk you through them one by one. I'm gonna make a video for each one because as we add more and more nodes, we can go ahead and create specific training videos for each node. This is not going to be showing you how to actually build an agent A to Z. We are going to have other training videos on that. This is literally going to be walking you through each individual node one by one. So without further ado, let's go through the Milestone node. So if I click the Milestone node, you can drag like that, or you can just click and it will be added on. Another way you can actually do that is by hovering this, you see this little blue button, then we click that and it automatically connects the Milestone node. And the Milestone node specifically is like a point in time of conversation. It's a specific milestone that we need to hit and achieve. And the specific reason that we have this little icon here is because if you're walking in the forest or walking down the path, just like any conversation, these are specific points in time where you can actually, you know, have different paths. You can go path A, path B, path C, which you'll see in just a second. But the Milestone is going to be one of the easiest to build with and also one of the most like predictable in flows and how it conversates and how it responds. And that is because of our setup here. So a Milestone node is specifically made up of a goal. So if you see right here, it's we tell you, 'Tell your AI agent what goal it should accomplish. Keep it simple and short as your AI agent performs better with short, easy to understand goals.' You know I've often had times where people come in here they're like, 'Tell the lead that we are XYZ and see how they feel and dig deep into their pain points and all this bullshit.' That is not how you leverage the goal. The goal is simple, clear, easy to understand, right? Things like, 'Determine if they're open to scheduling an appointment' or 'Get their phone number.' So if we build out a quick little flow here together, I'm going to say 'at' actually I'm gonna say 'determine if we're speaking with' and then instead of hitting this at button to insert a variable, I can actually just come in, I can use my keyboard and just put you know shift two to type the at symbol and then it'll put here so I can do first name. Determine if we're speaking with first name. Then I can give it extra instructions here. I can do things like say this in your first message, right? I can tell it what to say. 'Hey Alisha here.' I like to always do smiley face emojis. That's why I like to use women's names because it's kind of weird if men are always throwing up blushy smiley emojis. Um so we can come in, we can actually say break. If you type the word break like this with this closed and open arrow, it will tell the AI it needs to send this in multiple messages so it'll actually break it apart. So I always like to do that. 'Hey Alisha here, we're gonna send that as one message, break. Just to confirm, this is @First Name... Right?' And also just to kind of read what we put here in the tool tips. So the extra instructions is going to be extra help or information about to tell your AI about this goal. Use this to give more details or tell it exactly what to say. So for example, if it's an appointment booking node, the example is, 'This appointment is regarding our AI sales agent audit. We will audit their sales pipeline and build a strategy to launch an AI sales agent. If possible, include 'My boss Matty wanted me to reach out and offer you a free AI sales audit. Can I send you some availability?' So we can actually leverage this not just to tell it what to say, but we can also give it more context about this specific point in time. So that is what the extra instructions are for. And then we have our output flows. So the output flows is going to tell your AI what responses are good enough to move forward. Your AI checks if any of these conditions have been met. If one matches, the conversation moves to the next step. If none match, the AI stays here and keeps trying. So for example, if this is determine if we're speaking with first name, I'll say yes, we are speaking with the right person. Again, like it's just conversational, right? Like don't overthink it, what do I say, what do I say? Like don't overthink it. My favorite way of explaining this is explain it as if it's talking to a 10 year old, right? Simple, sweet, like you don't gotta get crazy and drop a freaking paragraph. like simple, very easy to understand, right? So we'll go yes we're speaking with the right person. And then here we will say, 'no we are not speak' or 'no the user denies that we are speaking with the right person.' Right? So they denied it's not me. Um so this would be two different options. Now one caveat to mention, I did this on purpose, I only left yes or no. So the AI is specifically looking for a yes where the person is going to say, 'Yes, I am Matty, how can I help?' or 'Yes, it is me.' So a a sort of confirmation that we are speaking with the right person, or if the user denies where they say, 'No, I'm not Matty,' 'No, you got the wrong number.' So it's only these two options that allow it to move forward. However, if the person says something like, 'Sorry, Alisha from where?', question mark. So it neither denies nor agrees that it's the right person. The AI is not going to move forward. It's going to stay inside of this specific node which is this milestone, and it's going to answer their question, overcome their objection, whatever they threw at us, and then it's going to re-attempt to get the goal accomplished. So if it says, if you said something like, sorry Alisha who? then the user neither confirmed nor denied, so it'll actually go to the next one. It's not just going to say my boss, Matty, it'll actually like conversate and it'll flow the conversation between the two nodes very nicely. So if I said, 'Sorry, Alisha who?' it'll say, 'Ah, I apologize, I'm actually Alisha from Agent Kong. My boss, Matty, actually wanted me to personally reach out.' That's why I said if possible incorporate this in your message because it's going to take the response from here and incorporate it and try to achieve the goal forward. So that is why I did the user neither confirmed nor denied, push it there. And then if the user denies that we're speaking with the right person, maybe we just want to kind of stop the agent and actually this would be the agent complete and done. I promise you if you tried this very simple agent, you would shit yourself as to how powerful this is. I know it looks simple and basic, but I promise you something like this will book a lot of appointments and we see it right now actively. We build a lot of these small simple agents and they are crushing it. But that is at the end of the day how you use a milestone node. You give it a goal, if you need to, give it some instructions and then define what allowed it to go forward. Actually, I will also put this one here. The user neither confirmed nor denied that way, just to kind of give you the example of how this would flow. If we say, 'Hey Alisha here, just to confirm this is Matty, right?' and I'm like, 'Sorry, Alisha who?' then the user neither confirmed nor denied, so it'll actually go to the next one. It's not just gonna say, 'My boss Matty,' it'll actually like conversate and it'll flow the conversation between the two nodes very nicely. So if I said, 'Sorry Alisha who?', it'll say, 'Ah I apologize, I'm actually Alisha from Agent Kong. My boss Matty actually wanted me to personally reach out.' That's why I said if possible, incorporate this into your message because it's going to take the response from here and incorporate it and try to achieve the goal forward. So that is why I did the user neither confirmed nor denied, push it there. And then if the user denies that we're speaking with the right person, maybe we just want to kind of stop the agent and actually this would be the agent complete and done. I promise you if you try this very simple agent, you would shit yourself as to how powerful this is. I know it looks simple and basic, but I promise you something like this will book a lot of appointments and we see it right now actively. We build a lot of these small simple agents and they are crushing it. But that is really at the end of the day how you use a milestone. You give it a goal, if you need to give it some instructions, and then define what allowed it to move forward. Actually, I will also put this one here, the user neither confirmed nor denied that way just to kind of give you the example of how this would flow. If we say 'Hey Alisha here, just to confirm this is @First Name, right?' and I'm like 'Sorry, Alisha who?' then the user neither confirmed nor denied, so it'll actually go to the next one. It's not just gonna say, 'My boss Matty,' it will actually like conversate and it'll flow the conversation between the two nodes very nicely. So if I said, 'Sorry Alisha who?', it'll say, 'Ah I apologize, I'm actually Alisha from Agent Kong. My boss Matty actually wanted me to personally reach out.' That's why I said if possible, incorporate this in your message because it's going to take the response from here and incorporate it and try to achieve the goal forward. So that is why I did the user neither confirmed nor denied, push it there. And then if the user denies that we're speaking with the right person, maybe we just wanna kind of stop the agent and actually this would be the agent complete and done. I promise you if you try this very simple agent, you'd shit yourself as to how powerful this is. I know it looks simple and basic but I promise you something like this will book a lot of appointments and we see it right now actively. We've built a lot of these small, simple agents and they are crushing it. But that is the end of the day how you use the Milestone Node. And again, if you do have any more questions, please reach out. We'd be more than happy to help. If not, then I'm going to move on into the next video. One more thing to actually add on this Milestone Node. I completely forgot to go over our advanced settings. So, to show you what the advanced settings does, we're gonna hit edit, we're then gonna click on this. And as you see, we now have this option that says fields to sync and skip step. So, if we don't have any fields to sync, you can't skip the step. What basically that does is it will sync this field. So we have name, email, first name, last name, phone, company, address, city, state, zip code, country. And if we choose the phone number, for example, which is exactly what we did here, get the user's phone number, it will then take the phone number and update the contact's phone number for you. So this is actually how you can generate leads with Agent Kong is by asking for specific details and then syncing it to your CRM. So if you use GoHighLevel or if you use our platform internally, it will sync it to that. If you use HubSpot, then again it'll sync it to your HubSpot. But this is the syncing happens, which is really really cool. In order to do that, you're going to open advanced settings, you're going to come down here, you're going to choose the thing that you want to sync, and if you don't want the AI to ask this because you already have it, you can hit skip this step. And if we already have a phone number, it's going to completely ignore this question and move right on to the booking node. So that is one other thing to mention is you can actually leverage the Milestone node to sync fields to the contact by selecting the right fields to sync and then hitting save.

All right, welcome to the training on the conversational trigger node. Now the conversational trigger node is quite literally exactly what it sounds like. It triggers specific flows at any point of conversation. So how we're using it in this agent, for example, this is a booking confirmation and rescheduling agent.

So here we're using a conversation node. Someone just booked a call, again, contact the lead to book a call with us, greet them, tell them to watch this video, then tell them if you need to reschedule your call just a minute, I can help. Right. So it's just a booking agent. And if at any point in time, the person sends a message and says, Hey, I need to reschedule my appointment, the AI knows that we need to trigger this because of the description.

So the description is basically the prompt again, this conversational trigger allows you to trigger new flows mid-conversation based on specific user input. Perfect for scenarios like I want to reschedule my appointment, I need to cancel, they have other questions, whatever, they can go ahead and trigger that.

So yeah, honestly, it's nothing too crazy. Once you understand the concept that you can trigger a new flow at any point in conversation, it allows you to get really creative. This is one example, right? The user is in here. They send a message. They say, hey, I'm looking to reschedule. The AI understands that we have a schedule, a trigger here that says the user is looking to reschedule. So we'll go there or to say, hey, I need to cancel, then this will pick it up because the user is looking to cancel their appointment.

It will then go in and then it will ask them which appointment they have and which one to cancel if they have one. Uh, and yeah, very, very simple there. Let's actually pull up another example just so you can start to see more of how this would work.

So here we have a flow. This is one that we built out on one of these trainings. We want to see if we're talking to the right person, then we want to see if we can offer them an available time slot and book them in for an appointment. But let's say throughout any of this time, let's say maybe the lead becomes hostile, aggressive, or irritated/angry.

Like maybe there's an issue. Maybe the guy says whatever, maybe he's frustrated. Actually, we looked into a conversation, and someone was talking to a lead that they already actually had as a client. And I guess the lead got into the system again, they rebooked, became a lead, and then the AI reached out and they were pissed off because no one reached out and they thought the AI was a real human and like, oh, you only reach out to new leads. I've been trying to reach you for a while.

Again, business problems happen across the board. Not my place to talk about or comment about that. But this could actually be a perfect example because we did see someone become hostile. So we actually came back and added this in there after the fact, just because, you know, at the time we didn't realize that people might be irritated or hostile.

But this, for example, if the AI is triggering at any point in time, maybe talking about a booking, and the guy, for whatever reason, is upset. Maybe he's asking how much it costs and you trained the AI not to talk about pricing, and he keeps pushing for it. And the AI keeps not answering it. And maybe he gets mad and starts swearing or like gets irritated, aggravated.

The AI would then realize this after every single message. So the way this actually works, important caveat, after every single message that the AI receives at any point in time, like it could be up in here, it could be in the booking notice. But every time the AI receives a message, the first thing it does is it scans to see if we have any conversational triggers that fit the conversation up to this point.

So if the lead is starting to become hostile, aggressive, and seems irritated or angry, the AI will see this as an option, it'll then match this to the message and the conversation. And if this conversational trigger fits, it'll actually trigger this conversational trigger right here and then obviously we're going to stop the agent because we don't need to keep talking, the person's probably irritated as we keep responding, let's just kill the conversation and let a human jump in to save the day.

So this is another example again, this could be triggered at any time. At any point throughout the conversation, it's like, yo, the guy's becoming irritated, hostile, aggressive, let's stop the node, maybe we want to create a new one here, conversational trigger. We've actually seen this one quite a few times, the user no longer wants to talk to us / is not interested.

Right, maybe the guy's like, hey, not interested, please stop messaging me, or hey, stop messaging me. If this is not there, depending on where the conversation is at and what type of output flows you have, it may still follow up. But if we have this here, you know, the user no longer wants to talk to us / is not interested, then if at any point in time the AI receives a message and it thinks that this is a good fit, it'll send it here, and then it'll push it down to the next node, which is now going to be the stop node.

So this is another example of how to use the conversation nodes, because again, these conversation nodes can be triggered at any point in time for the conversation, and it just gives you way more control, way more fine-tuning on how you should handle conversations.

And yeah, that is the conversational trigger in a nutshell. If you guys do have any questions, feel free to reach out to us on Discord or to the support team. And I'll catch you guys in the next video.

All right, so here is the training video for the send message node. So this is the message node. As you see, very simple one sends a message and continues to the next node. If we go ahead, we can hit this quick add button. Bang, connect to the send message node, and let's get into it. Again, fairly simple. You got two things that you can do with it. You can either prompt or static. And it's quite literally what it says. Static is going to say, hey, Matei's a G. It'll say that literally word for word, letter for letter. Perfect. Everything. Hey, Matei's a G.

Um, if you know, although I don't think it'd really make too much sense. Actually, no, you could use like, hey, at first name, Yeah, let's do that. Hey, at first name, did you know Matei is a G? Uh, and by the way, if to do this, I'm just holding shift number two. So we hit the at, we want to do this with our keyboard. If we do that, then it'll pop up the variables. You could then search name first name, all your custom fields are going to be in here as well. Uh, which is actually where I would leverage this a bit more. So you can again, say a static message, it'll say letter for letter, word for word, but if you go into the prompt here is where it gets a bit more interesting.

Here is where you can leverage, uh, so it's a prompt, obviously, it's going to the AI is going to right here, uh, the prompt is going to ask the AI to create the message based off your instructions, and then static has been to send the exact text that you write. Uh, so the prompt, you know, you could say the lead is spending, I don't know. Uh, I think I have a monthly ad spend in here. Cool. Monthly ad spend and is generating. I think I saw something about leads. Cool. How many leads you generate per month, per month.

Uh, use this data and tell them a positive compliment. I'm just kind of riffing off the top. I think it's kind of shitty at night, but maybe at some point you want to compliment someone. But this is again, it's going to take the lead is spending, let's say $1,000 a month and is generating 50 leads per month, use this data to tell them a compliment. So the AI when it hits this node, it'll be like, Hey, I see that you're spending whatever and you're generating whatever. That's awesome stats, very excited and they can move forward. Um, I'm honestly not that big a fan of starting with send message notes. I actually don't use send message note all that much.

Mainly when I'll use a send message note is if I'm somewhere here, right, like, you know, the user denies that we're the wrong person. Maybe I want to say something like little message, bring this down here, bring this over here. By the way, visually, I don't know, I just, I found kind of like the hack to creating visual, uh, flows. I kind of just like to line them up in rows and it just kind of looks very nice, clean and organized, very rare random side note.

Uh, but here's maybe where I'd use a send message note where it's like, okay, apologize for disturbing them and wish them a great day. So you could do that in the prompt, you could do static. This is really where I would, where I would go ahead and use the send message node because you know, we're about to end a message. Cool. I'm gonna, I'm gonna send one last thing. Uh, I'm not a big, big fan of starting with a send message. I'm not really using it too, too much. There's the odd time where it makes sense. And I think this is something that it would make sense, right. Like right before you end right before you do something. Uh, but yeah, that's a send message node. Again, fairly simple. You can prompt, you can use static, it sends a message and once it sends the message, it goes on straight to the next. It's not like the milestone node where it'll send a message and kind of stay here until it finds an output flow that it matches on a condition. It literally just comes in, does what it needs to do and then continues forward.

All right, welcome to the training on the conversational trigger node. Now the conversational trigger node is quite literally sounds like it triggers specific flows at any point of conversation. So how we're using it in this agent, for example, this is a booking confirmation and rescheduling agent. So here we're using a conversation node, someone just booked a call with us, greet them, tell them to watch this video, then tell them if you need to reschedule your call, just let me know and I can help.

Right? So it's just a booking agent and if at any point in time, the person sends a message and says, hey, I need to reschedule my appointment, the AI knows that we need to trigger this because of the description. So the description is basically the prompt and this conversational trigger allows you to trigger new flows mid conversation based on specific user input. Perfect for scenarios like I want to reschedule my appointment, I need to cancel, they have other questions, whatever, uh they can go ahead and trigger that.

So, yeah, honestly, it's nothing too, too crazy once you understand the concept that you can trigger a new flow at any point in conversation. It allows you to get really creative. Uh, this is one example, right? The user is in here, they send a message, they say, Hey, I'm looking to reschedule. The AI understands that we have a trigger here that says the user looking to reschedule, will go there or if they say, hey, I need to cancel, then this will pick it up because the user is looking to cancel their appointment, it will then go in and then it will ask them which appointment they have and which one to cancel if they have one.

And yeah, very, very simple there. Let's actually pull up another example just so you can start to see more of how this would work. So here we have a flow. This is one that we built out on one of these trainings where it's very simple. We want to see if we're talking to the right person, then we want to see if we can offer them an available time slot and book them in for an appointment. But let's say throughout any of this time, uh, let's say maybe the lead becomes hostile, aggressive or seems irritated/angry, right?

Like maybe maybe there's an issue. Maybe the guy says whatever, maybe he's frustrated. Actually, we looked into a conversation um, and someone was talking to a lead that they already actually had a client and I guess the lead got into the, they rebooked, became a lead and then the AI reached out and they were pissed off because no one reached out and they thought the AI was a real human and like, oh, you only reach out to new leads. I've been trying to reach you for a while. Again, this is problems happens across the board, not my place to talk about or comment about that. Uh but this could actually be a perfect example because we did see someone become hostile, so we actually came back and added this in there after the fact, just because you know, at the time we didn't realize that people might be irritated or hostile, but this, for example, if the AI is triggering at any point in time, maybe it's talking about a booking, uh, and the guy for whatever reason is upset, maybe he's asking how much it costs and you trained the AI not to talk about pricing, uh, and he keeps pushing for it and the AI keeps not answering and maybe he gets mad and starts swearing or like gets, you know, irritated, aggravated.

The AI would then realize this at after every, so the way this actually works, important caveat, after every single message that the AI receives at any point in time, like it could be up in here, it could be in the booking node, but every time the AI receives a message, the first thing it does is it scans to see if we have any conversational triggers that fit the conversation up to this point. So if the lead is starting to become hostile, aggressive and seems irritated or angry, the AI will see this as an option and it'll then match this to the message and the conversation and if this conversational trigger fits, it'll actually trigger this conversational trigger right here and then obviously we're going to stop the agent because we don't need to keep talking. The person's probably irritated as going to be responding. Let's just kill the conversation and let a human jump in to save the day. So this is another example. Again, this could be triggered at any point throughout the conversation. It's like, yo, the guy is becoming irritated, hostile, aggressive, let's stop the node, maybe we want to create a new one here, conversational trigger. We've actually seen this one quite a few times.

The user no longer wants to talk to us/ is not interested, right? Maybe the guy's like, Hey, I'm not interested, please stop messaging me or hey, stop messaging me. Um if this is not there, depending on where the conversation is at and what type of output flows you have, it may still follow up. Uh but if we have this here, you know, the user no longer wants to talk to us/ is not interested, then if at any point in time the user or the AI receives a message and it thinks that this is a good fit, it'll send it here and then it'll push it down to the next node, which is now going to be the stop node. So this is another example of how to use the conversation nodes because again, these conversation nodes can be triggered at any point in time for the conversation and it just gives you way more control, way more fine tuning on how you should handle conversations. and yeah, that is the conversational trigger in a nutshell. If you guys do have any questions, feel free to reach out to in the Discord or to the support team and I'll catch you guys in the next video.

So this is going to be the training for the wait node and the wait node is quite literally just like it says wait x amount of time and move on to the next node. So you can wait 5 seconds, you can wait minutes, hours, days. And we're not really using this a whole whole lot because once we get into the settings training, you'll see that you can have a uh where did it go? You can have a message delay. So the AI will wait 25 seconds before generating a response. This can be 120 seconds, it can be however long you want it to be. Uh so we do have already a setting for waiting in between messages. So it's not like you need to send it or add it in between every single node. That would just be redundant, annoying and ugly.

It's very much if you want specific fine tuned control. So let me show you actually where we would use something like this. So right out the gate, one of the most prominent ways to use this is if you have multi-channel architecture. Now by multi-channel architecture, I mean simply this, you have, you know, different channels and each channel is going to fire the second this trigger goes off. So once the agent gets triggered, it will send all three conversations in all three channels. So if you want to prioritize, maybe we'll do SMS first. So it triggers it goes through, we're not going to use any weight steps, but maybe after that it's going to go to what's up.

So if we click on what's up and we maybe delete this one, we kind of pop this back a bit and we add a wait step. It will wait, maybe two minutes. Boom. And then it will move forward. Now we're adding a bit of control, right? So we have iMessage, it goes first. We don't have any wait step in here. Now WhatsApp has a two minute wait step and then email, I don't think I set up the email flow. Uh maybe it might just be actually for a test. Let's not pay attention to the actual use case for the email because this is probably not what you would do inside of an email.

Um but let's say we had this properly set up, we would probably want to come in here at a wait node, wait, and maybe here we'll wait 3 minutes. Then and then we'll send the first email. So this then basically gives you a controlled approach as to when each channel is firing because again, this one has no weight step. WhatsApp had a two minute weight step and then email had a three minute weight step. So first it'll send the text, then it will send the WhatsApp, then it will send the email.

I'm actually thinking about this, the AI is probably gonna send a message regardless on all. So maybe instead of waiting two minutes, we'll wait 1 minute for this because we'll send a text in 60 seconds that he'll get hit on WhatsApp and then wherever they respond first, the agent will pick up. Um, so that is probably how I would do it is like, you know, zero, one minute, maybe two minutes or 1.5 minutes, but that is one use case to actually leveraging the wait node. You could also leverage this in between nodes specifically. So, you know, again, we have our 25 second message delay here and maybe for whatever reason before it goes here to the apology and farewell, maybe you want to add a wait step. I don't know why. Um, again, just kind of letting your brain and your mind understand how this could work. But maybe here we want to wait, you know, two minutes. So once the person responds, let's say he says, no, it's not me at the wrong number, the AI is going to wait 25 seconds to see if he sends any more messages once the 25 seconds time delay hits, then it will pass forward and then it's going to hit the wait note, it'll wait two minutes and then it will send this message. So again, it just gives you more fine grained control over maybe that odd case scenario, uh where you do want to add a wait step for whatever reason, primarily, I would say the most common use case is going to be for controlling which flow starts first. So it feels more natural and they don't get an SMS or WhatsApp and an email all in one feels kind of spammy, right? It's SMS first, then a minute later, WhatsApp, then a minute later, email. Uh So that is probably the most common use case, but again, you can use it at any point in time and once it hits it, again, it will wait two minutes and then it will pass it on to the next node, which in this case would say apologize for disturbing, wishing them a great day. And then boom, it will stop the agent. So that is how you use a wait node. Again, fairly simple. So I'll catch you guys in the next video.

All right, so welcome to this training video. We're going to talk about the stop node and the stop node is honestly is exactly what you think it is. We hover over this, you're going to see it says stops the agent from talking to the lead across all apps and unassigns the agent. So once we get to the stop node, it will stop the agent completely from talking not only across WhatsApp, but from SMS and from email because WhatsApp is not the agent. iMessage is not the agent, email is not the agent. This is the agent, right? The agent is everything.

So once it hits the stop node, it basically turns the agent off from the contact completely. So if the contact ever replies, the agent will not engage with them at all. So that's pretty much it. Uh if you want to maybe make a funny analogy, imagine this as a birth canal and you have a load of swimmies going to find the egg, which is, you know, in this case booking the appointment. That's like the golden juice, right? Booking the appointment or disqualifying themselves.

Uh so it's like, you know, these are really just trying to get there to one of these nodes as fast as possible. And once something enters into it, boom, it won the game. No other opportunity is available. No one else can hit the stop node. It will completely disengage the agent, completely from that human, from that lead, from that contact, which is yeah, the end game, you want the agent to do its job, complete its job and hit a stop node. So I'm not going to beat around it anymore. I think this is a very simple and easy to understand node. You use this when you want to stop the agent completely.

One important caveat to mention if you do have follow-up settings, which we'll get to in another module, which is going to be dialing in your agent settings. But let's say you do have, you know, some follow-ups where it'll continuously follow up and let's say instead of stopping the agent for whatever reason, you add them to a conversation node, right? We'll just call it test.

If you add it to a conversation node, if you still have follow ups that need to happen based off your follow up settings, it will actually continue to follow up with them. So let's say they say they're not interested in booking the appointment and you send them here or what I've seen other people do which is actually more realistic with use an actual use case. Someone books the appointment and then they want to keep the conversation alive. The AI will say uh keep the conversation alive. So if it does the booking and it goes here, if there are any follow ups on this channel, which is I think WhatsApp or if there's any follow ups on SMS or email, again, it will actually continue to follow up with them. It's not going to stop the agent. So that is why we recommend once the agent has completed its job, send it to a stop node. That way it can wrap it up. It's finished, it's done and it's not going to continue following up with the lead. But that's it for this video. I'll catch you guys in the next one.

All right, so welcome to the training video where we go over the booking node. The booking node can also be found right here. You can just drag it on the canvas, or, oops, you can click on the palette to add it to the screen, whatever your preference is. But the booking node is a simple node. The booking node, once it enters, it starts to immediately try booking. Now, there's a few caveats here.

The first thing is the confirm time zone. So if confirm time zone is on, the first thing that the AI does when it hits this booking node is it says, what is your time zone? It'll confirm the time zone of the contact. And then I would always recommend save contact time zone being turned on as well, because then we'll update the contact inside of your CRM so that it saves with the proper time zone.

And why that's really, really cool is when this is turned on, the AI will ask for the time zone. And then if I say I'm in Dubai, it will understand what time zone that is and it will give me the time availability in my time zone, not the time zone of the calendar. Then when I select the right time zone, the AI will take that time and go book it back in the calendar in the time zone of the calendar. So it's not, you know, trying to book 8:30 Dubai time in 8:30 EST, it'll do the calculation, it'll revert it back to the calendar and then it will book it accordingly, which is really, really cool.

So if confirm time zone is on first, again, once the AI pushes things out and it goes to the booking node, the first thing that it will say is awesome, what time zone are you in or what time zone works for you, something along the of that nature is the first message that gets sent in. If this is turned off, so if we hit edit and turn these off, and that's how it stays, then the AI will provide them availability. So we have trained Agent Kong from the ground up to be an appointment setter. So it's not, you know, a GPT wrapper or an 8N template, everything is custom built from the ground up. And we specifically trained it how to be an appointment setter. And one of the things that we trained it to do is when it grabs availability, it's always going to provide the next two available days on the calendar. So if today is Thursday and you take only Monday to Friday, it will offer Friday and Monday. If today is Saturday, it will offer Monday and Tuesday if that's the next availability, so on and so forth. I think you understand.

So that's what an appointment setter is trained and is supposed to do, right? Like anyone who's built appointment setting teams, they know, hey, go grab the next two available days and provide them dates. So that's what it does if confirm time zone is off. Once it hits this, it'll say something like I have, uh, tomorrow the 21st available or next Monday, which one works for you. So that is if this is turned off, if confirm time zone is turned on, which is what we like to do, especially for international booking, then it'll ask which is your time zone and then it will go and it will provide the next two available dates from the calendar that is selected.

Now, with that being said, that means that you need to use this kind of strategically, right? Because if you think about conversation, if you just out of nowhere are like, I have tomorrow or Thursday and you've never spoken about booking or anything, it's kind of strange, right? So there's a few ways that we do this. Uh number one, this is actually an Instagram DM or messenger DM, uh determine if they're open to scheduling a live demo, uh the user stated they are interested in scheduling an appointment, then we get the phone number to obviously have the phone number to book them in so we can call them, et cetera.

Uh so they've already agreed to it, right? It's like, awesome, are you, can I send you some availability? They'll say yes. Then the AI will say, perfect, real quick, what's the best phone number for you? They'll give them the phone number and then it'll go awesome. What time zone works for you or if it's turned off after he gives the numbers like, awesome, I can do tomorrow or Friday, which one works for you. So, important caveat, make sure you're setting up the booking appointment so that it makes sense conversationally by asking them if we can set them availability or if they'd be open to booking a call. Uh so that way they're not just getting hit out of left field because if you had, you know, the right person and then you go into the booking node, be like, hey, is this Matei? and be like, yeah, who's this? It's like, awesome, I can do tomorrow or Thursday.

Uh so make sure you set up the booking node. It's really all the point is, make sure you're setting it up. I'm actually going to find another example here to show you. Perfect. Here's another flow. I think we actually built this one together on one of the other ones, but very simple, determine if we're speaking with the right person, ask them if we can send them a call, be straight to the point of why we're reaching out if they say yes, then it'll be like awesome, I can do tomorrow or Friday because we don't have confirm time zone turned on. So always make sure that you set up the booking appointment so that it makes sense conversationally. So the person has said yes, cool, send me some availability and then it will go ahead and do it.

All right, so now that we've gone through the regular use case of the booking to actually book appointments, it's very important to mention that that's not all the booking node can do. The booking node can also reschedule appointments. So this is going to be the workflow that you can leverage something like this. You're going to send it into a conversation node to keep it open-ended that way. The AI is always on without pursuing any specific, specific goals or feeling as rigid. And then if the user ever says at any point, I need to reschedule my appointment and you have a conversational trigger on the screen, then it will send it down into the booking node, which if you see right here, is how we are going to actually reschedule the appointment. So it's important to mention that this is not going to create a new appointment inside of your calendar. It's actually going to find the original appointment and then edit and modify the date and time of that appointment, quite literally rescheduling it in your calendar, not creating a new one.

So that is something that's really cool. If by the way people have multiple appointments, which obviously does happen if you've run any online ads, you've probably seen leads that have two appointments booked on different dates. So like, all right, which one are we keeping? If that happens, the AI will see, hey, I see you're booked on Tuesday at 3:00 p.m. and Friday at 4:00 p.m. which one would you like to reschedule? Um and then you can all it will also ask like, you know, should I cancel the other one if it thinks it's necessary. Uh so the AI is smart, the booking node is not only for creating new appointments in the calendar, it can also be used to actually reschedule the existing one or if you see right here, if the user is looking to cancel their appointment, we can send it into the booking node and the AI is smart, it understands exactly what's going on in the conversation. It knows what's been spoken up to this point in time and it will go ahead and it will ask them if there's multiple, which one do you want to cancel, if there's obviously not multiple, just cancel the one that's inside of the calendar, but it will actually cancel the appointment as well. And important to mention that by cancel, it's not going to delete the appointment, it's just going to mark it as canceled. So it no longer shows up in the calendar and takes up the space. So that is the complete tutorial on the booking. Again, you can use it to actually book appointments or you can use it to reschedule appointments and cancel appointments. So it does three functions and it's all inside of one nice beautiful node. Now, if you do have any more questions, of course, feel free to reach out to the support team. We're more than happy to help and assist.

All right, so on this training video we're going to be going over the tag node, which basically allows you to edit or delete tags. So as you see, right here, we can go ahead, we can drop this here, and when we drop it, we will then be able to add a tag or remove a tag from the contact and then we can obviously select the tags that we're going to remove or select the tags that we are going to add. So very simple, you either add or remove a tag. Now why this is actually important and why would we want to actually use this? Well, for example, if you have a third-party CRM like HubSpot or GoHighLevel, whenever you add a tag to a contact in our system, we automatically send that data over to your third-party CRM because there's always a two-way sync going on between any CRM that you have connected and our platform. And on top of that, if you leverage something like let's say GoHighLevel, you can actually trigger an automation once a certain tag gets added, again, in an external system. So just to show you an actual use case, this is for a car dealership seeing if we can actually or if they're okay with us calling them. Um so if they say yes, they would like us to call them, then we will uh add a tag here that says call and then that way inside of HighLevel, it'll trigger an automation sending the sales rep an instant notification to call the team, or if it's only text, so please text me, then they will add a tag saying text only, and then again that's going to trigger a separate automation inside of GoHighLevel so that the sales rep gets a text only notification. Now, obviously, that is just a very basic format, you can definitely use this to trigger automations for very complex stuff. I'll let your imagination take you wild. But again, in this training video, very simple, you can always go ahead, open the node palette, find your Edit Tags, drop one on, and then you can simply add tag or remove tag and then select the tag or multiple tags that you would like to add. And that's it for this training video. If you guys have any questions, please feel free to reach out to the support team. If not, then we'll see you in the next video.