Transcript
Danchez: There was a time when I was too successful as a marketer, which is like, how many times do you get those in your career, right? I was generating hundreds of leads every single week, and it was so good that my sales department just couldn't keep up.
Danchez: I was working for a college, we were growing enrollment, and it was crazy how many leads I was able to generate. It was the time when Facebook was was like the the leads were just so cheap. If you were plugging into Facebook ads early on in 2016, 2017, and it was a major problem.
Danchez: If you generate too many leads too fast and you can't get with those leads on the phone, call them up, follow up, as as good as the eat automated email sequences I had were, it still wasn't enough to convert all those leads.
Danchez: I wish... we had the level of AI that we have today back then, because I would have been able to not just 3X the enrollment of that university, I would have been able to 10X it with the amount of leads that I was getting and couldn't follow up with. It'd be like weeks later.
Danchez: And that's why I'm excited to have this conversation today. Today, I'm excited to announce, I'm having an interview with Fur Patel, who is the director of special projects at HighLevel, but you're also like the AI guy within HighLevel, because we're goingnna be talking about conversational marketing, or at least the features that high level has that has conversational marketing in it.
Danchez: I'm excited because like the problems that I used to have with not being able to follow up with leads, not being able to give them a number to call where they can actually find a human to answer his questions and book meetings. Those days are gone and it's changing the game. I am so excited to be having it first. Thank you for joining me on this podcast.
Fer Patel: Oh, my pleasure. I'm excited to be here. I'd like to share more about what AI and what we can do at high level to solve those pain points that you brought up.
Danchez: Man, and they are pain points for lots of businesses, all kinds of businesses all over the place. In fact, I was just talking to a friend of mine who's a chiropractor and he's like dying because like he wants to answer the phone.
Danchez: People call in. They don't want to, not all of them want to book online. They call in, it's missed call. Why? Because he's cracking some backs, man. He can't be answering the phone call while he's like adjusting, cracking someone's neck, man. That's like, you can't be doing multitasking, that stuff.
Fer Patel: Yeah.
Danchez: So, but beyond local businesses, I know it's a big deal for all kinds of businesses out there. So before we get into the conversational AI, for I'd love to know a little bit more about your background. Like before you were into high level, like what were you doing and then what what led led you getting into using this this software? Yeah.
Fer Patel: Yeah, okay, that's an interesting story. I'll go ah <unk> <unk> have a short version of it. So I was in the fintech space, financial technology space, and I did a lifestyle pivot and moved into the marketing world about six, seven, eight years ago, I believe.
Fer Patel: And when I got into the marketing world, I was using various marketing platforms, a funnel builder platform, a email platform, then Zapier and Google Sheets and chatbot platforms.
Fer Patel: I must have had about six to nine different platforms in my arsenal or my toolkit at any point of time for my clients. And as you know, if you're using Zapier and other solutions to kind of glue it together, you're hoping you know it works great.
Fer Patel: But when it doesn't work, everything breaks down at the same time.
Danchez: Yeah. Yeah. yeah
Fer Patel: And then you've got to figure out where the problem is. So a lot of automation issues. I'm paying. six to nine different invoices every single month. You got to maintain all them. You hit different limits for every single platform.
Fer Patel: So there's various issues they have to deal with. And this is at the same time, High Level came out, launched in its infancy stage during that time. And I saw it the first time. I'm like, okay, another marketing platform.
Fer Patel: Great. I don't have time to learn another marketing platform.
Danchez: Yeah.
Fer Patel: It's a conundrum. You know, do you take a stop, stop now and focus on something to make it accelerate you or no time, no time, and just focus on what you're doing. So I, you know, I ran into that same issue, no time.
Fer Patel: I kicked it down and wrote about two, three months. And then at that point, like, all right, got access to the account. I started playing with it. I got into it. I saw the value in it right away. I'm like, one platform that can consolidate my six to nine various tools within one that can automate everything in one centralized area. I just have to log into one platform every single day.
Fer Patel: It was a game changer. And I was one of the early adopters, I would guess sub 5,000 in the Facebook group, but one of the early adopters in the community. and And I've never looked back ever since I've been the high level space ever since then as a agency owner and as a user before becoming part of high level.
Fer Patel: So this is where my journey began as an agency owner. I then provided aftermarket tools for other agency owners, how to customize their dashboards, how to improve the user experience. so if you want to call it hacks, how to just create a better user experience while high levels building out new feature sets.
Fer Patel: I was creating add-ons around it in my own way, providing aftermarket tools. So I had the opportunity to interface with probably thousands of early adopter agencies in the ecosystem over the year.
Danchez: Yeah. I remember my journey really started with Infusionsoft. I mean, before that was MailChimp and a lot of us were using MailChimp and starting to integrate the API. Even like I was doing this in like 2010 and even integrating MailChimp into at text message marketing so people could text into a short code and then it would kick into the email system and follow have a drip sequence. This was really advanced for like 2010. Yeah.
Fer Patel: Yeah. Oh,
Danchez: But then I heard Infusionsoft build this journey builder and it was beautiful. And I, it's well, Infusionsoft was like a small business software. I used it for college and built the whole college admissions process on it.
Danchez: And like, we were crushing it with Facebook ads.
Danchez: They were going into Infusionsoft and they didn't do phone calls and text messaging back then. had to get a separate thing called fix your funnel, really robust add on for Infusionsoft.
Danchez: The man, I had that thing humming to the point where other colleges were like, what are you freaking doing? And I'd be like, well, we're using this thing called Infusion. It's Facebook ads, mostly in Google ads.
Danchez: We're sending the landing pages and it's Infusionsoft. And then it opened up the campaign. you know You'd see the little drag and drop builder and all the different campaigns we'd be running. And they'd look at me with bright, big eyes and be like, how much?
Danchez: I'm like, what do you mean how much? I'm showing you how to do it. and they're like, no, do it for us.
Fer Patel: Yeah.
Danchez: And I'm like, I don't have time to duplicate this. That's why when I first discovered high level, because of the big difference, it was like, I'm like, it's Infusionsoft with Fix Your Funnel baked tank because it could do the phone call texting thing, except you can duplicate the whole thing over to it another sub account.
Fer Patel: With non-shotting.
Danchez: and having limited sub accounts. I was like, mind blown. I was like, oh my gosh, I wish I would have had this. I have made so much money with all those colleges coming to be out. But like, oh yeah, I can resell that. I could just duplicate the whole thing over for you for X amount a month.
Fer Patel: yeah and You know, it's funny they mention that.
Danchez: I'm like, oh my gosh.
Fer Patel: Snapshots is one of the things that if you were an agency owner running multiple clients within the same niche, it's the snapshot feature is what really pulled me in because I was onboarding, for example, a real estate client.
Danchez: Yeah.
Danchez: Yeah. Yeah. Yeah.
Fer Patel: And they have tons of emails and the drip sequences. Then you gotta pipe, you know, what type of client are there? Are they, you gotta really work it through. And then once you get locked down, then you onboard a new real estate client, rinse, repeat. It takes about 10 hours of onboarding, copying and pasting. It was a nightmare.
Fer Patel: The snapshots, 20 minutes, change some custom values in your live. Now you just got to focus on double down on that niche. It's easy.
Danchez: yeah
Fer Patel: You have' it get one working perfectly the way you want it and then maybe make some quick little edits, but it's no longer a tedious onboarding task. And that was one of the things that drew me into a high level when I came on board.
Danchez: And that's why it's such a it's such a great tool for agencies, but even consultants and freelancers. like We're all making recommendations on how to best build the funnel, and we have our preferred way of doing it. But I'm like, it's so dirt cheap.
Danchez: I remember I was paying more for Infusionsoft. I was probably paying Infusionsoft and Fix Your Funnel together, $1,200 month. And I was at a nonprofit license like discount, so it was probably like $2,500 worth.
Fer Patel: OK. OK. okay
Danchez: like, for ah ah for high level at like a month, or if you're doing one, like stupid cheap. I'm like, my gosh.
Fer Patel: yeah
Danchez: So that's it from just a software perspective. But now it does a lot. It does. There's so many tools in here. a big part of it, of course, is AI being infused into it. Before we get into the AI features, like were you using AI before you started working at high level? Was that like a normal part of your your workflow?
Fer Patel: It wasn't actually. i wasn't really using AI in terms of my day-to-day, except for the obviously chat GPT.
Danchez: Yeah.
Fer Patel: And that's the standard now. It's just and what we all know. When you think of AI, you think of chat GPT because they're the ones came first to market.
Danchez: Yeah. Yeah.
Fer Patel: And They're the early ones. Now you have numerous other players out there, many other alternatives. Now I use a mix of different tools or depends on my use case. If it's more content heavy, I'll use this tool.
Fer Patel: If it's more coding heavy, I'll use another tool. So depending on the use case, but I wasn't using AI per se much in terms of a business utility before I joined high level.
Fer Patel: But this is at the same time I joined High Level was when they were launching the Voice AI product. So it was launching it at Level Up in October 2024. joined high Level literally but a week before they already launched it.
Fer Patel: So I came in, was already being launched.
Danchez: Yeah. Yeah.
Fer Patel: So I had nothing to do with the initial inception of the product and go to market.
Danchez: yeah
Fer Patel: So this is where I joined on the AI product. And it's the background I bring is being from the agency side, understanding the user experience, understanding all the pain points as a agency, the ups and downs, getting the clients and losing the clients and all the churn.
Fer Patel: So this is where i come in. And what what a lot of times I taught when I'm talking to a team internally, I tell them, My role is to be the voice of the agency owners. I know, you know, I don't represent all them, but i have a good idea of what agency owners want and what they're going to like and not like.
Fer Patel: So this is where when we're rolling out products in the AI space, when a lot of times you're a team will be building, they're focusing, they're working on the features, they're rolling it out. But then my question is, like how how is the agency owner actually going to use it?
Fer Patel: It's a great feature, but how are they going to use it? What is a their experience going to be like?
Danchez: Yeah.
Fer Patel: How are they going to adopt it? And then once they adopt it, how are they going to leverage this tool to generate more revenue for their agency? A tool is great, but the whole idea of High Level is we're giving you a lot of different resources and tools to empower you to make more money.
Fer Patel: You have a one-shop, all-in-one CRM. You have AI, but we want you to use these tools for your own agency. That's a given, but we want you to resell these tools for your clients so they can benefit from the same solutions that you're benefiting from as well.
Danchez: the one thing I want to specify is that it's a fantastic tool for agencies, but individuals can come and use this tool as well at the $97 a month plan. But even at the $97, you get access to three sub accounts. So you could like run two businesses out of it and still have one for your friend or ah client.
Fer Patel: Well, that's kind of the idea. like I think the way the team envisioned it is you get three accounts. One account is for you and your lead generation for your usage. And the other two is for you to onboard your first one or two clients.
Fer Patel: And hopefully if you onboard i a a single client, you break even, then you onboard a second client, you're in profit.
Danchez: Yeah. Yep.
Fer Patel: So therefore the first two clients pay off your main expense of the high level. And now, Your goal is to grow and get more and more clients and the software becomes a source of revenue generation. It's not an expense center for you.
Danchez: Yeah. And while a lot of agencies love it because you can kind of perfect one account for a specific industry and then just duplicate that over and over again, because like if you're doing it for like a chiropractor, like my friend, like it's kind of the same sequences over and over again, because most chiropractors have the same business model. So you can duplicate it over and over again.
Fer Patel: Yep.
Danchez: But even for an individual, all the tools are there that you would ever want.
Fer Patel: Agreed.
Danchez: like from Calendly, funnels, automation, texting, phone calls. And it's they're not even like lightweight. They're like robust. Like I go back into HubSpot sometimes. I'm like, it can't do X. It can't do all these things.
Danchez: It even does the more advanced things that HubSpot does. Like- You build custom objects in there. I'm like, oh my gosh, that's like some sophisticated stuff. The fact that you can build in your own custom objects. And if you're not familiar with what that is don't worry. It's just, it's it's an advanced CRM feature.
Danchez: can build your own databases in there of information. But it can do all these things, but it's really individual. for It's really good for almost any small to mid-sized business. Like I can't, like HubSpot's so expensive now that I can't imagine anybody using HubSpot until they're making like at least 10 million.
Danchez: and revenue.
Fer Patel: Okay, wow.
Danchez: I'm like, I don't, I don't know who, who would, who would benefit from it unless maybe you're like heavy enterprise B2B sales from the get. Maybe. Yeah.
Fer Patel: Yeah, I've used it in the past when I had my own startup. And the reason I was able to get into HubSpot as an early user is you get a startup discount at 90% off the first year.
Danchez: Yeah.
Fer Patel: And I think it's 50% off the second year. And then by the third year, you're paying full rack rate. At that point, we canceled out because we couldn't afford it anymore.
Danchez: Yeah.
Fer Patel: It was just wait. Because then we kind of grew into all the different little add-ons.
Danchez: yep yep.
Fer Patel: So we were paying a couple of hundred bucks a month. Then all of a sudden by year three, it's skyrocketing up to thousands of dollars a month now. like guess it's we're not we weren't We did not see the justification of paying that much money for a CRM application where we can easily find alternatives at a lower cost that can do equal or greater at that point.
Danchez: One of the exciting things that love about High Level is this really robust conversational marketing suite. And that's where I want to spend the rest of the time. So tell me, like, give me an overview of like what High Level's conversational marketing tools are like. Like how many channels are there?
Danchez: Without AI, we'll go into the AI of them all because that's really where they're starting to shine is into the AI part.
Fer Patel: yeah
Danchez: But like what what are the different channels it has for conversational marketing?
Fer Patel: So conversationally, you know a lot of people, when they come on board, the fundamental obviously is email. Everyone does email and then SMS. Those are your two core. I think SMS high level really scaled SMS as a ah ah solution, as an as an offering across all marketing agencies. Because before to get SMS wasn't a singular solution.
Fer Patel: You would have to go to a third party dedicated just for SMS.
Danchez: Yeah. Yeah. Yep. yeah
Fer Patel: Then you would have to use a solution just for email. Like I don't remember ever using SMS until coming up to high level because I didn't want to buy a third party solution just for SMS alone. So now you have your core email SMS and that's what everyone started out of the gate using high level.
Fer Patel: Those are your two options. Now you also have Facebook Messenger integrated within high level. You have Instagram Messenger integrated within high level. We also have WhatsApp. So WhatsApp is one of those things I find quite interesting.
Fer Patel: it's It's not really adopted much in the US, WhatsApp as a whole.
Danchez: I know it's super international, but Americans don't use it. I have to explain this to my Canadian and UK friends all the time.
Fer Patel: Yeah, so I'm surprised.
Danchez: Like we don't use WhatsApp here.
Fer Patel: i
Danchez: Yeah.
Fer Patel: use but I WhatsApp.
Danchez: Yeah.
Fer Patel: I'm in Canada. I use WhatsApp all the time.
Danchez: yeah
Fer Patel: And the reason I use WhatsApp all the time is because I'm communicating with people outside of Canada. Family in UK, people in India, people in different parts of the world. WhatsApp is the way to communicate over Messenger. It's simple.
Fer Patel: If I need to make a call, it's free. So there's no issues over there. So I've been using WhatsApp for many years, but at the time, communicate with people in the US. like i don't I had an account. I don't use it because they don't have a purpose for it.
Danchez: It
Fer Patel: So WhatsApp is a massive... tool outside of north america because a lot of them don't use sms sms have heavy costs in a lot of those regions whereas whatsapp no cost right or unless you pay a dollar um mean i don't even know what they i think they were that one dollar a yearly license
Danchez: used to be a dollar a year. Yeah.
Fer Patel: Yeah.
Danchez: I don't know what it is now.
Fer Patel: So we have a WhatsApp. We have SMS, email, Facebook, um um Instagram. Those are your core. Plus, we have live chat widgets you can put on your website. So there's a lot of different ways of bringing people into your world.
Fer Patel: But if you want to extend the concept of communication, it's like the whole either idea of capture. How are we capturing people coming into our world? And we can extend it out with our forms, our surveys, because the main starting point is how do we get leads into the engine?
Fer Patel: How do we get leads into high level, the whole capturing process? What are all the different ways it can come into our world? Oh, and the other thing as well, by extension with that is If someone calls the phone number, that's another way of bringing them into your ecosystem.
Danchez: Yeah. Yeah.
Fer Patel: They all become contacts. Once you build up a whole contact base based on the communication channels, now you have multiple ways of engaging them or nurturing them. So we want to keep nurturing them because we're not capturing we're not in the business of just capturing contacts the sake of capturing contacts.
Fer Patel: We want to capture them to have a conversion event. What is that conversion event?
Danchez: yeah
Fer Patel: Either they forced you know get them to make an appointment, get them to make a purchase, get them to opt into a webinar. What is that one conversion event? It all depends on the use cases. And you mentioned they can do so many different things with high level. you know The way you know we talk about sometimes is high level is like a marketing operating system.
Fer Patel: You can do whatever you want with it.
Danchez: Yes.
Fer Patel: You have so much choice. You can go any different direction you want with it. It's really up to you. So you can keep it as simple as possible or you can make it as complex as possible. It's all up to you and that's why sometimes a lot of people prefer snapshots because someone's already taking that thought process away from them.
Fer Patel: Here's a snapshot going to start with. So I have my emails, templates, I have my SMS templates, i already have the content that works for them. I have my landing page with a form. That's all I need. I throw that up, then start driving traffic to that page and get the whole process going.
Danchez: And it's funny, like complexity is really the enemy. It's trying to keep it simple, which can be hard because if you learn all the tools instantly, you want to start making it complex. And really it's like, it's like part of the learning journey is like, first, you know, nothing and you're struggling.
Danchez: Then you learn a lot and it becomes too complex. And then you become a master. And usually by that time, your, your campaigns are really simple.
Fer Patel: to simplify
Danchez: Your landing pages are really simple. It's like, but that's how it goes.
Fer Patel: it. We've seen cases where people are They have, you're trying they're trying to find the optimal SMS message. What's a perfect message.
Danchez: Yeah. Yeah. Yeah.
Fer Patel: And they're just struggling to get it. Then you have those who have been doing it for a while. You have some people in the space. have someone called Rob, someone named Rob Bailey, who's an early adopter of high level. And he's coined the whole database reactivation campaigns. And he just shared with us, you know, what are you writing your text messages?
Fer Patel: It's like, it's literally a line and a half. Hey, we got this offer. Are you interested? Yes or no? And that's it. And then once if they respond, yes, then the whole process begins. But we make a whole people in general, people want to put as much information as possible on that first text message.
Fer Patel: And people look at us too much. They don't know how, they don't know how to respond back with a yes or no. It becomes complicated.
Danchez: There's something I want to say to the audience, and it's because we're talking about AI so much, I rarely ever talk about the other marketing channels, but I could i could do whole shows on every like most of the marketing channels. There is a complete overemphasis on email when it comes to marketing.
Danchez: And if you're listening to this now, chances are you found me probably through like Social Media Examiner or LinkedIn. You probably come maybe more from a B2B space where the huge emphasis is on email.
Danchez: I'm just telling you the biggest opportunity is not an email. It's in text messaging. Like everybody pays attention to it. There's hardly a filter there. and I've, when I grew the university was pretty much on the back of text messaging.
Danchez: Now that's a student audience.
Fer Patel: What's that?
Danchez: So like, of course they're not even in the realm of like, you know, if once you're in the office space, you're more into email, but outside of that text messaging of the thing. And if you're international, then, What was it?
Danchez: the That social app is is the primary thing.
Fer Patel: but that
Danchez: WhatsApp is the thing. But text messaging is not utilized hardly enough by marketers. And high level, was kind of like one of the first big things that it was like it doubled down on.
Danchez: Before, i even Infusionsoft didn't do texting at all. You had to go use Fix Your Funnel. And Fix Your Funnel was early in being able to do the thing where you can automate the blasts that go out. And at any time you could take over so it became conversational, your actual sales reps, as soon as they replied to the automated thing that went out after they became a lead, they could take over the conversation, which was just freaking magic at the time because no no software did it.
Fer Patel: Yeah.
Fer Patel: yep
Danchez: Fix Your Funnel was one of the first to actually make it reasonably priced to actually have your call center actually chat with people via text. High level came in and did the same thing, but it was actually baked into the system. Unlike Infusionsoft, it was a third party. High level is like all in one.
Danchez: You could actually do that kind of work.
Fer Patel: Yeah.
Danchez: But now it's even more exciting because now you don't even have to have your call reps be there to answer any anymore.
Fer Patel: if
Danchez: Because that was the problem I had is I had all that sophistication in the software where like people would fill out a form and then it would send them a text message 15 minutes later being like, hey, saw that you just asked for information.
Danchez: what What was exciting about it? Just to kick off the conversation. you know And they respond and then they wouldn't hear back for a while because maybe they were staying up late and it was like 11 p.m. and couldn't man the phones that long. you know It's hard to man all the weird hours. a hard man hard to have people over the weekend. you know It's a lot of hours you have to put. It's 168 or how many how many hours we have in a week? it's hundred and sixty two 162, something like that.
Danchez: That's a lot of hours to man, especially if you're working 40-hour work weeks. But now that's not a problem because... AI can answer it, which is what I want to get into next.
Fer Patel: yeah
Danchez: So how is high level incorporating AI into this stuff?
Fer Patel: Yeah. Let me just add on to that one last point here. and That's a great example you just brought up that you just reminded me because I used to work with an agency in the car dealership space.
Fer Patel: And they need to help automating all the different inbound channels through all the different lead capture sources. So cars.com, Edmonds, and a few, you know, 910 others.
Fer Patel: Because emails come in, or it just sits there, no one responds back because they're sitting in someone's even email inbox. So we turned every email, converted it into an SMS message outbound. We pulled it in, then send an SMS to the client saying, hey, we got your we got your inquiry, your interest. you know let me Tell me more if we want to just warm up the lead.
Fer Patel: Now the issue is they responded back. And now the sales rep is busy. He's you know doing a demo. He's out, takes you know two, three hours, maybe someone respond back. So then we came into that same situation where someone got, you got them to engage and then nothing happened.
Fer Patel: Then you, they moved on.
Danchez: Yeah. Yeah.
Fer Patel: So now with AI, This is where the beauty is. So if we want to take a step back over here, so look at that situation here.
Danchez: yeah
Fer Patel: We initially send an outbound message to engage the client and we just need them to respond back. Once they respond back, the AI solution, conversation AI will take over.
Fer Patel: Because based on the prompt that you set up, based on the agent you set up, and the whole goal is, okay, let's get you in for a test drive, for example.
Danchez: yeah
Fer Patel: That is the core focus. Someone responds back, like hey, I'm interested in this car. Perfect. Okay, what what what color are you looking for? what is this? Do you have a trade-in? When are you looking to change? You know just yeah ask your pre-qualifying questions.
Fer Patel: No problem. you know Are you interested in setting up an appointment for a test drive? that's the goal. That's the focus of the objective of the bot. Yes. When would you like to book it? you know This week or next week?
Fer Patel: Well, this week is good. Okay, perfect. Morning or afternoon? Oh, morning. All got availabilities on this day, this day, this day. which way Which day works best for you? The person will select it.
Fer Patel: Perfect. Okay. It's booked. Looking forward to seeing you. you have any questions, feel free to reach out to this number. And the AI did the whole process without the salesperson intervening to ask the exact same questions.
Fer Patel: They'll come in. They'll meet with a salesperson because now that that's where the human touch comes in for the conversion. But that whole process of booking the appointment, the conversation AI part or the chat bot took over and handled it perfectly because that is exactly what it's set up to do.
Fer Patel: This is where we would normally lose leads because we forgot about them. it came in after hours. We lost them. They're no longer interested. And if you're a aggressive buyer, you're putting posts out.
Fer Patel: It happened to me. I was buying a car. I messaged three different companies same time. The first person responded back to me.
Danchez: the person who calls back first.
Fer Patel: They got back to me and said, what when do you want to come in?
Danchez: It's huge.
Fer Patel: Can I come in today or tomorrow? They go, today is perfect. I'm on my way. And I literally bought the car within six hour window from query because it made sense.
Danchez: Yeah. Yeah.
Fer Patel: You know, obviously all the other conditions made sense, but the first one that responded back to me and it worked within what I was looking for, I sealed the deal right away.
Danchez: Yep.
Fer Patel: And the other two called me back. I think one call, it was on a Thursday. One call me back on a Monday. The other one called me back a Wednesday after. like Sorry, I already bought a car. It's too late. You lost your chance.
Danchez: Thank you.
Fer Patel: So this is the idea where all the AI tools, if you leverage it well, it's not even complicated.
Danchez: i
Fer Patel: set it up. It's the whole core purpose of that. In this case here is to a book a test drive, which is technically book an appointment.
Danchez: So many things. I just had the same thing happen. I was calling about a bunch of local businesses and I was just going through the list. I was opening up Google Maps, started making phone calls. Actually, i was i wasn't buying something. I was selling something, but it's something they all wanted.
Danchez: I was liquidating some silver because the price of silver went up high right now. i was like, oh, i got a little stash. I got a box of silver, junk silver coins. So I just started calling all the coin shops. The guy who called me back first was the guy I ended up selling all this stuff to.
Danchez: He got everything. ever Other people called me back or some people didn't even, never answer the phone, never call back. I'm like, no, the guy who was the most responsive to me and started texting me, he got the deal.
Fer Patel: Yeah.
Danchez: He got, he, he won. So it's a big deal. Like we've all heard the stat of like, you know, like I forgot what the conversion rate is, but there's like a massive advantage to what if you can call someone back within five minutes.
Danchez: And we've all experienced it when we get called back within like, especially within 10 minutes of filling out a form.
Fer Patel: ye Yeah.
Danchez: It almost feels like magic because it's so, it happens so seldomly, right? Someone gets you and be like, hey, I just saw that you were looking on the website and you were asking for information.
Fer Patel: Yeah.
Danchez: You're like, I was. When it's within 15 minutes, guess what?
Fer Patel: Yeah. Yeah.
Danchez: You have it fresh in your mind. You have not forgotten.
Fer Patel: yeah
Danchez: Because I don't know, I've worked the phones before and it is kind of embarrassing when you're calling back like weeks later and they're like, who is this again? Nah, I don't think I gave you my information. You know they did, but they even remember.
Fer Patel: yeah
Danchez: It's embarrassing.
Fer Patel: I've experienced that with local service businesses.
Danchez: Yeah. yeah
Fer Patel: I was trying, i was calling people for roofing related and call them, left a voicemail. Then I emailed them and like, what's the point sending emails? They never respond back to emails. They rarely call back unless they're that busy.
Fer Patel: and they don't need any more business, someone is dropping the ball significantly. And all they need to do is put an automation in there. because all I want to do is just get me a quote or get me an appointment.
Fer Patel: I don't want talk to anyone specifically in general. I just want to know, answer my question. Can you do this? Yes or no? How much will cost? And if it works, I'm booking it. That's all I wanted. But I'm always amazed by you go to website, fill out a form, and no one responds back you. Well, what was the point of that form?
Fer Patel: So for anyone listening, any client who has a form that you fill out and they don't respond back their clients, you can go after to convert them to AI.
Danchez: It's opportunity.
Fer Patel: They are the one of the perfect clients and go after. They're not responsive. It's a gap.
Danchez: So let's break this down as far as to what this actually looks like right now. How many channels can you implement this like AI answering questions, booking meetings kind of thing through this AI agent?
Fer Patel: So the number one way people use it is live chat. On the live chat widget, you put it on your website.
Danchez: Yep.
Fer Patel: So clients come to your website and they can interact and they can go through the regular process.
Danchez: Yep.
Danchez: Yep.
Fer Patel: That's a standard process. Then you can take it offline and you can do it through SMS. So someone texts in, you text a number, hey, I just saw you called or you're interested in this, you opted in, you SMS.
Danchez: Someone texts in. Yep.
Fer Patel: Are you interested booking employment? Yeah. And now the AI takes over the whole SMS communication.
Danchez: ah be huge. Even if you're speaking from stage, imagine, because a lot of speakers do that, right? They're scan this code. Sometimes they do it right there. Sometimes it could be like, text this number, text this word to this number. All of a sudden you can have an AI agent working the whole crowd while you're still speaking.
Fer Patel: 100%, yeah. SMS, live chat. You also have Facebook Messenger. So some people are communicating with you through Facebook DMs online your or through messengers there. AI will take over.
Fer Patel: Instagram, AI will take over as well. We have as well, and WhatsApp. So WhatsApp recently in the last couple of months. Now WhatsApp is also part of your AI toolkit for communication channels.
Fer Patel: So now the the only one we're not doing yet is email. Email is a little more, it's not, it's just, you know, it's one those, we can do it, but we haven't done it yet.
Danchez: Yeah. Yeah.
Fer Patel: You can do it through our workflows. You can figure it out yourself. But the reason with emails is it's just, it's little more convoluted to create a better response.
Danchez: yeah
Fer Patel: And this is more for quick conversational, email's not conversational per se. It's more, it's the structured communication back. Whereas WhatsApp, one or two lines, are you looking for? You want to book an appointment? What is it? It's very conversion focused.
Fer Patel: Right.
Danchez: So Instagram's the one social channel. Are there other social channels that are coming on the line? Like, are you ever at LinkedIn?
Fer Patel: We are exploring LinkedIn still.
Danchez: LinkedIn's one of those weird ones because I think it's against terms of service when it comes to LinkedIn.
Fer Patel: We're just a
Danchez: But yet everybody does it.
Fer Patel: link' a little more complicated.
Danchez: ah ah LinkedIn just doesn't do anything about it.
Fer Patel: everyone has their own Everyone has their own hacks around it until they get shut down, then they switch it up.
Danchez: Yeah, yeah.
Fer Patel: So we're trying to go through the official route with LinkedIn through the API teams and all that.
Danchez: Uh-huh.
Fer Patel: So it's just been a long communication journey.
Danchez: Yeah. Yeah. That'll be exciting. Cause I mean, there's a, there's a, don't There's considerate ways to automate these things and there's inconsiderate ways.
Danchez: Like I would never want an AI pretending to be me and me faking it.
Fer Patel: Yeah.
Danchez: Like it's like, Hey, Dan, Dan here. I'm like, no, it's going to be, Hey, this is Dan's AI assistant. Thanks for reaching out. Um, can get a hold, I can book something with you and Dan. Um, is there any questions I can answer first before I connect you with Dan?
Danchez: You know, like that's can just considerate. It's just giving them a way to get their questions answered and, and get the thing that they want, but without pretending to be me.
Fer Patel: so you know
Danchez: Which is why how I'd be implementing it on something like social.
Fer Patel: Yeah, so it varies. you know Some people, depends if they lock down their AI perfectly and it works out well for them on Facebook, Instagram, and even SMS.
Fer Patel: They don't even tell people that it's an AI. They just handle themselves because they're not going to find out generally unless the AI goes off completely, right?
Danchez: Yeah.
Fer Patel: But in the most part, it works. However, so that one is just like, it depends how you what your...
Danchez: Yeah.
Fer Patel: perception is how you want to go about doing it. It's up to you in terms of preference.
Danchez: yeah. yeah.
Fer Patel: Whereas voice ai it's one of those that we I advise people, let everyone know it's a voice AI assistant because no one wants to make be made you as if they were fooled on this call.
Danchez: yeah
Fer Patel: Because you're a one minute into the call, everything sounds great. Then all of sudden, this sounds like an AI. Then you feel foolish as if you're talking to a, you thought you were talking to a human. Now you feel a little bit foolish and you kind of feel frustrated. Whereas chat, you know it doesn't come across. You can get away with chat, but voice AI, voice AI to a certain point,
Fer Patel: You know, you can get away with it, but at a point, you feel if people feel they're being deceived, theyre not going be happy about the experience. So ah ah initially, i would always say, hey, thanks for calling.
Danchez: Yeah. Yep.
Fer Patel: I'm Alex, this person's voice AI assistant, or i'm ah I'm a voice AI, or I'm an AI. I'm here to help you answer your questions. Just massage it in. People don't mind. it And from experience, from the voice product itself, is people don't mind talking to a voice AI when you're calling to get things resolved for you.
Fer Patel: Right? We've had seen great adoption on the inbound side. If I'm calling, it's an AI that answers and it's able to answer my questions and take and achieve the goal I want. Hey, I want to book an appointment and it does it for me.
Fer Patel: I'm happy. I don't have to talk to someone. i don't have to waste any time. I achieved my goal. What people hate and do not want is an AI agent calling you to sell you something.
Danchez: Yeah, nobody wants this.
Fer Patel: Right? People, there's a large demand in the community as a whole.
Danchez: Yeah.
Fer Patel: Everyone's like, oh I want outbound AI. want to be able to call. I want to cold calling. But then we ask them, we take pause. Great. Sounds great in theory. Sales side. Take a step back and put yourself in the client's shoes.
Fer Patel: Now you're getting that AI phone call. Do you want an AI bot calling you anytime trying to pitch you on a product? like No, I don't. Like, go. Right?
Danchez: Don't do it. Yeah.
Fer Patel: like Don't do it. It's not a good. So this is where we've been slowly, you know, we're not, we haven't walked into the whole outbound side. and We see other people are doing it. were like We're taking our time on this one here.
Danchez: Yeah.
Fer Patel: And it has to be, it has to fit specific use cases.
Danchez: Yeah. Yeah.
Fer Patel: And our goal is to be very much conversion focused outbound. So even conversion focused outbound, what I mean by that is someone is calling to remind you about your appointment.
Danchez: yeah
Fer Patel: And hey, you have an appointment coming up. do You have any questions I can answer for you. and Oh, yeah. Am I supposed to fast before or not? Great question. This is what you need to do here. i'm going to send you an SMS with all the details that is solving a pain point.
Fer Patel: People don't mind that.
Danchez: yeah
Fer Patel: Or hey, just let you know, we have a webinar coming up at 3 You know, it'd be great if you're there 10 minutes early because of XYZ. If you have any questions and feel you make sure you got the link. if you don't have the link let me know. I'll send the link right now.
Fer Patel: Again, you're solving a problem conversion focus. You're giving a warm touch. Someone already knows the relationship. Whereas if you do a cold outreach AI call, people hate it.
Danchez: They've never heard of you. Ouch.
Fer Patel: If I get a call and it initially sounds like someone's talking to me, I try to talk them.
Danchez: Yeah.
Fer Patel: Then they all of a sudden i feel like, wait, this is AI. I hang up.
Danchez: Yeah.
Fer Patel: Everyone's going to hang up.
Danchez: Yep.
Fer Patel: That's just a natural.
Danchez: mean, they hang up on humans. They're going to be even faster to hang up on your Yeah.
Fer Patel: This is where, you know, from the high, low perspective, we tread very carefully outbound because regulations, compliance, there's laws coming effect.
Danchez: Yeah.
Fer Patel: People are getting fined quite a bit.
Danchez: Yeah.
Fer Patel: We see other players in the market. if People tell me, well, they're, you know, they have it. They have it. but Yeah. They also can find and find and find. You don't hear about that, but we hear about these common communications. So outbound will come about in the near future for us, but very specific use cases in a very compliant way.
Danchez: Like workflow triggered outbound to where they're already in the workflow. I guess you could still hack that to just feed it a cold list and it would start outbounding people. But there's got to be a way to do it.
Fer Patel: Yeah, we have ways we have ways of mitigating it.
Danchez: Yeah.
Fer Patel: We have processes in place. we're just you We're just focusing on the inbound experience, making that stellar, improving that. Then outbound is just going to be a switch for us.
Danchez: Yeah, I'm excited because I've seen you. You even have like a grayed out button of like, you have it set for inbound, but I'm like, there's a grayed out button for like outbound. I'm like, ooh, it's coming. and And that the kind of situations you're talking about are the ones I want to use. Like someone fills out a form and then you follow up with the phone call that asks, hey, I got this, is what's coming up. You filled out this thing. You're getting the X. I just want to check to see if you have any questions.
Danchez: If not, then, you know, we can part ways, you know, simple things like that.
Fer Patel: Or a perfect, yeah, exactly, perfect example. Someone opts into a Facebook form.
Danchez: Yeah.
Fer Patel: You run a campaign. you're interested, to let me know. Then now you want to that warm touch, call them. Hey, I received your inquiry about you're interested in buying a house. I'm so-and-so's personal AI assistant.
Fer Patel: I just have a few quick questions to make sure I get you the right information that you need. You're like, sure, no problem, because I'm interested in buying a house.
Danchez: Yeah. Ugh.
Fer Patel: I did reach out to you, and now you're reaching out to me to ask me some qualifying questions to get me what I need. I don't mind. But getting an AI caller, are you interested in selling your house?
Fer Patel: Like, who's calling me?
Danchez: oh
Fer Patel: Like, no.
Danchez: If you want to see what this voice sounds like, and they just recently dropped some new voices that are much better. Like it's getting better all the time, but the new voices are sounding really good now. have I set up an example for this morning for all everybody listening. If you want to see an example of what this sounds like, and you want to book some time with me to talk about how I'm using high level, that's what this phone call, I don't know, AI voice agents trained to do. It's a train to book meetings with me to talk about high level with you and just answer questions about high level.
Danchez: I actually did a chat GPT deep research to find and scrape the internet for the most frequently asked questions about high level. I went and put it into the FAQ section of the the voice agent. So it has the most frequently asked questions.
Danchez: You might ask it something that it doesn't know and I'll see the call script and then I'll be able to and upload that into it. um if you want to test it out, 615-610-0721 is the phone number and go and chat with it. I think you'll be pleasantly surprised how fluid it is, how good it is. It's not perfect. It still messes up.
Danchez: But you I mean, High Level just released this less than a year ago. And it's already gone through multiple iterations of improvement, which for those who aren't in high level already, like this is kind of like high levels MO.
Danchez: Like they tend to release things rough and then let the community give feedback and then they improve it quickly.
Fer Patel: That is...
Danchez: Usually it's like they launch three months later, there's like the fixes and then another three months go by and then there's the 2.0 version. It's kind of like high levels rhythm, not always on that timeline, but generally that's how they go.
Danchez: They launch a whole new thing, they fix it and then they improve it.
Fer Patel: Well, the MO is more referred to as a skateboard model. We release it with a skateboard to eventually get to a car.
Danchez: Yeah. Rapid iteration. Yeah.
Fer Patel: And the iterations are not every three months. It's every week to every two weeks.
Danchez: Yeah.
Fer Patel: It's content continuous updates until it gets to a good, stable production-level product that people are able to use. And we've been seeing great success with people, with clients over the last many months already.
Danchez: So I want to break it down farther because I think people, when they hear AI agents, I've been talking a lot on the show about AI agents just not being there. like Like most of the time when someone's saying the word AI, the agent thing, they're they're just blowing smoke.
Fer Patel: Thank you.
Danchez: It's not actually useful. This is one of those few examples other than like ChatGPT's deep research research. This voice AI thing, and all your conversational AI, but particularly the voice, is definitely agentic because you can load it with information.
Danchez: It can have tools available to it, like multiple tools I noticed in there recently. I don't know what to do beyond like calendar booking, and it's smooth. it's not It doesn't even just send you a link. It's like, oh, yeah, we have some availability on Thursday afternoon and on Friday morning.
Danchez: Or is 10 a.m. m on Friday good? or should we look at Thursday afternoon? It actually like conversationally books meetings, which is amazing.
Fer Patel: yeah
Danchez: If you call the number I gave earlier, you'll find out. But it can do more than that. It can actually like take actions. It can launch work automated workflows. It can send text messages. And I'm like all the tools it has available to it at its disposal, depending on where it feels like the conversation is going. I'm like, this is 100% agent and it's powerful. It's cool.
Danchez: What else?
Fer Patel: Yes.
Danchez: What else are you seeing people do with it?
Fer Patel: So, you know, the different actions you can do is a a more ah ah ah call transfer. Someone says, hey, I want to but want to talk to your manager or i'm interested in sales or I want to purchase.
Fer Patel: So you define what the criteria is before the call gets transferred. So you build out the prompt saying, hey, if the caller is interested in making a purchase or buying or shows they're interested in buying now, transfer call to specific phone number.
Fer Patel: So we have a call transfer functionality. And before it gets transferred over, they'll say, OK, I see you're interested in making a purchase. One moment, please. And I'll transfer you to someone in the sales team.
Fer Patel: And then they'll do that as a call transfer. So call transfer one action. The other action is that we can send an SMS real time on the call. So I've seen people who are using it for their restaurant. They kind of do a pre- pre-reservation or pre-ordering bot they put it in place.
Fer Patel: answer it here I'm here to answer any questions you have before you're ready to place the order.
Danchez: Yeah. Yeah.
Fer Patel: So whenever you're place an order, to say, hey, want to talk to a human and then I'll transfer you over. So they ask, okay, how so you know what how big is your large pizza? What are the top specials of the day? They can take their time because when you're talking to a human on the phone at a restaurant, you feel like you're rushed. You got to book it or the guy's like, look, you know i got I got to deal with all the other people.
Fer Patel: So They've been seeing great success on their side of people talking to the AI, asking all the questions, getting the menu items. Or someone says, oh, do you have a gluten-free menu or or allergen menu? Like, sure, no problem. i'm going to SMS you the link right now.
Fer Patel: And real time on the call, we're going to send them an SMS with the link for the menu related to gluten allergies or whatever reasons that you want real time to send the SMS. We have trigger a workflow on the call. So workflow, um if you know anything about high level, is the most powerful feature in there. You can do anything you want with our workflow.
Danchez: Yeah. It's like the journey builder, the automation, drag and drop builder. It's where all the automation happens.
Fer Patel: Yeah, so you can do whatever you want in that workflow too. If you just choose whatever the prompt is, the person says they're interested in joining your mailing list. Perfect. That means on the back end, create a little trigger action, add their email to the workflow, put them into a whole sequence.
Fer Patel: It's all done. Tag them. Whatever your workflow does, it can be triggered during the call. And as well as you can trigger a webhook. So those who aren those that's more meant for more advanced users. So on the call itself, I've seen people create, you know, just for proof concept, they create a appointment or a weather bot.
Fer Patel: They call in. They want to know, hey, what's the weather like in California or in San Francisco? At that point, it does a call out to the WeatherBot API, and responds back.
Danchez: Cheers.
Fer Patel: And they will tell you, oh the weather is 76 degrees Fahrenheit like in real time. so But this if people want to use this webhook concept a little further, they can post to their point of sale solutions.
Fer Patel: They can post to any third party applications that need to know about this information based on whatever the outcome of the call is. So there's lot of function on the voice product.
Danchez: it's
Fer Patel: And one of the other things that people find amazing is voice product. you know If you sell to a client, they never have to log into their high level account. Because after the call is completed, we'll send them an email summary with a transcript of the communication.
Danchez: Yeah.
Fer Patel: So now the a the owner the business owner can here see, oh, I got a call. And you see exactly what the conversation was about. And you see what the outcome summary was as well, which is an amazing selling point because they see in real time what their AI agent is doing for them. like What work are they accomplishing? Are they working or not working? You see in real time those emails.
Danchez: Yeah. And then i I've been getting them because I've been testing the phone number that I just gave a minute ago. And I've been seeing my own conversations come in. It's really handy because i'm like, oh okay, this is useful. Because what if there's a frequently asked question that you didn't account for?
Danchez: Well, you'll find out and then you'll add it and it'll start answering that question appropriately. Right. You can actually see what the whole conversation was. And of course, it gives you the summary and the highlights, but then you can switch over to the transcript if you need to.
Danchez: But this is a powerful tool because essentially anybody for a very low fee, anybody can have essentially a full-fledged what a human receptionist would be doing.
Fer Patel: yeah
Danchez: I can't. It's pretty much what any human receptionist would be doing, except it is working 365.
Fer Patel: Yeah, so you know just on that, we had a client who was in the roofing space. And they would drive leads to the roofing company, and they would use a third-party service, an answering service, to receive the inbound inquiry and say, okay, thank you very much.
Danchez: Yeah. Yeah. Yeah.
Fer Patel: Someone will get in contact with you. That's all they were doing. Nothing beyond that. Very basic because they don't know anything about your business. They're not trained up. They just do very basic name, phone number, email, when should we call you back?
Fer Patel: That's it So they got ahold of the voice AI product and they jumped in you know feet first right away. In the first week, they converted five roofing clients onto voice AI and they were charging them anywhere from setup fee, I think was a thousand to $1,500, just a setup.
Danchez: Yeah.
Fer Patel: set up and And then per month they were charging 797.
Fer Patel: Right. And the client was super happy. They're excited. now they don't have to use a third party service.
Danchez: Yeah. It's way cheaper. Yeah.
Fer Patel: And it's cheaper. It's predictable. They see everything going on. It's trained up. They can book appointments. They can ask qualifying questions. What type of roof do you have? Is it a slanted roof? Is it this type of roof?
Fer Patel: Is it an emergency service? Is it this? You can go through a whole qualification with your voice agent. And then if it's emergency, automatically SMS the owner.
Danchez: Yeah.
Fer Patel: Okay, we got a hot lead for you right now. Call this person right now.
Danchez: do you Can it be smart enough? I haven't gone this far with it, but it'd be really cool. can it If it's asking qualification questions, can it kind of then do like the just basic figuring of like, hey, my next recommendation is going to be this versus this.
Danchez: You know, like, you know, they're not actually a good fit for the core offer. I'm going to actually recommend that they go on the newsletter. It's like, hey, based on what you know, like most people your size would actually do better with this. You know, can it make recommendations at the end, even if if it's pretty simple?
Fer Patel: Yeah, it's all based on the prompt, right? Based on the prompt, you can say, you know, as you're progressing through it, however set up your prompt, asking these questions at certain point, saying, okay, based on what you've heard so far, assess if they're a fit for this or that, then ask them if they're fit for this, ask them these type of questions. If they're a fit for this, ask them these type of questions.
Fer Patel: So you would calibrate your prompt
Danchez: Yeah.
Fer Patel: in the journey itself. So you're going through a sequence and you're asking these type of questions, to qualify them. Then you are pretty much advising the AI agent in the prompt saying, make an assessment of where they fit and then lead the questions down accordingly. 100%.
Danchez: Interesting. I'll have to experiment with that. It's essentially, if you can build it in a custom GPT, you could probably build it in this voice agent. Though I will say setting up this voice agent was a lot easier than a lot of other AI work that I've done.
Fer Patel: and
Danchez: Like you could, it's as simple as really like setting it up. You can hook up databases to them. So you set up a voice agent, dedicate a number to it, and then you can hook it up to different databases. So you can have a database for one product or maybe a database for a different product set up a new one just for high-level questions and trained on high-level website and my AIautomationvault.com website. So it can answer questions about that.
Danchez: And so it's hooked up to it. And then I just kind of modified the prompt a little better, especially the leading line. I think mine, if you call in, like again, it's six one five six one zero zero seven two one If you call in, it has like a leading line of like, hey, I'm Dan's AI voice assistant.
Danchez: you like to book a complimentary session with Dan to talk about high level? Or is there anything I can answer for you regarding high level? You know, and it's like either it goes right to booking or it starts answering questions. But at the end of every one, I prompted it in such a way to be like, hey, at the end of every question, also remind them like, hey, can i answer more questions? Or would you like to book a session with Dan?
Danchez: It'd be nice if I could actually like tailor where they're at and actually lead them down different paths, but I haven't gotten that far yet.
Fer Patel: Yeah. it's it's
Danchez: But it wasn't that hard to set up. Like it took me all of 30, 40 minutes this morning. Like it wasn't that hard of a setup. It was pretty simple.
Fer Patel: And hopefully, it will be even simpler in the coming months as we roll out a lot more templates for you to use. So depending on what industry you're in, you can select which template or which agent you want to install within your account.
Fer Patel: So if you're in the roofing space, the chiro space, or or ah accounting space, You choose, you specify, we'll pre-build out your prompt for you. Then we'll pretty much will give you 80% what you need.
Fer Patel: Then you need to refine it down to make it your own or customize it for your own case or remove things that don't make sense for your business.
Danchez: To kind of wrap this up on our pre-interview, we talked a little bit about where you're going. I'm super excited about where it's at right now. Is there anything you could talk about to like what, what you're seeing now versus like kind of what you see generally this, where you see this going in the future?
Fer Patel: Yeah, so short term, we're looking at rolling out outbound in the limited capacity. We're also looking at you know releasing a lot more languages capabilities within the platform. I think we support about eight languages right now.
Fer Patel: We want to expand that out to
Fer Patel: 20, 30, whatever is capable within the infrastructure. Because language is a little bit tricky because anyone understanding the whole flow, it's not just about the language itself. It's taking the voice, turning it into text, taking the text and running it through AI.
Danchez: Yeah.
Fer Patel: AI then responds back. We take that text, converting it back into voice, and then make that voice you know turn into spoken voice using the different voices that we're using. So it's a few steps involved.
Fer Patel: so There's a bit of complexity. And to make sure all this is done within a rapid sequence, you no one's going to sit there, hey, when are you available?
Danchez: Yeah.
Fer Patel: And then just sitting there thinking, thinking, thinking for eight seconds. And it's ruining the experience.
Danchez: You missed it Yeah.
Fer Patel: It needs to be very fast, low latency. So we're improving on that side. We're also, you know when you're referring to databases, we're on our end, we call it knowledge bases.
Danchez: Yeah. Oh.
Fer Patel: We're expanding out how many different knowledge bases and what the knowledge base can do. So then with the knowledge base, it can We can have it structured in a way that if you're running an e-commerce site, you can pull up inventory on the call.
Fer Patel: Hey, do you have this in stock? It knows how to pull up the information within your knowledge base. Yeah, we have three in stock available right now. So this is, you know again, short-term improvements. Longer term, where we're going is
Danchez: Come on, dynamic knowledge bases. I don't think people realize like how complex that is.
Fer Patel: longer term.
Danchez: like Right now, you can feed it like FAQs, it's static. You like give it the question, the answer is you can train it on websites. There was something else you could put into it. I think spreadsheets were coming soon, but it's none of it's really dynamic.
Fer Patel: spreadsheets as well.
Danchez: So you're talking about actually having like dynamic, it can actually check dynamic databases, like inventory. That's cool.
Fer Patel: Yeah, hopefully. i won't i can't answer that question 100%.
Danchez: Eventually it's going to get there.
Fer Patel: But some
Danchez: When it's going to get there, who knows? That's the future though.
Fer Patel: and but some of the other things i'm working on is you know using the voice products of connecting to the real AI agents within high level. So we are rolling out or building out AI agents for all the various parts within the CRM.
Fer Patel: So you'll have an agent for funnels. You'll an agent for your CRM. You have an agent for workflows. You'll have an agent for emails. So you can then voice communicate using MCP within the infrastructure, calling the agents.
Danchez: Yeah.
Fer Patel: And from there, say, hey, need to send an email or I want to create a campaign to send emails to this smart list on this topic. Can you get this set up for me? and the agent will go and take care and build it out for you, then you have to go in there just finish it up.
Fer Patel: Or I want to build a funnel that does this. If you can communicate it to someone and it makes sense, the agent can go and do it because it has the same capabilities. So we're going we're leveraging the voice product that we have and making it easier within your high level account to give actions and execute actions at the same time.
Fer Patel: you can query again saying, OK, we're the most popular, and which email should I send or what's the most popular you know X. It gives you a lot more metadata that you want.
Fer Patel: And then hopefully, the not just the voice side, but on the AI side, we want to go from being reactive for the AI to proactively AI, when you log into your account, will come and tell you information about things you need to know versus you asking, for example, you need to say, you know,
Danchez: Yes. Yeah. Yeah.
Danchez: yeah
Fer Patel: You need to create a blog post. You need to create social media and social media posts. You don't do it. You forget. You move on. But now the agent will need to log in saying, hey, I've created three social media posts on the following topics. Here's what it is.
Fer Patel: you want me to publish it? Yes or no? so it's already done the job. Now all you need to do is say, yeah, proceed. Or i created a blog post. We're in the month of November. Here's a blog post for the month of December for Christmas or for Thanksgiving on topic with your brand, something you've done before. It's trained up.
Fer Patel: Here is a blog post you can go to market with. Again, it's not going to be 100% ready. 80%, 85%. So lot of times people are like, oh, I can't just publish it right away. like, come on.
Fer Patel: AI is not going to be replaced to us in 100% at this moment.
Danchez: Yeah, yeah, yeah.
Fer Patel: we have to re We have to just... temper our expectations saying AI will get you to 80%.
Danchez: We don't want to go 100% because sometimes it just goes off the rails. AI can get weird sometimes. in In these systems, like in a controlled system, like high level, it probably won't, but like peoples get people's chat GPT accounts get weird sometimes.
Danchez: I'm like, you don't really want it to go 100%. Not for a while. Like we really do need the controls to keep it from taking actions like sending emails or booking meetings sometimes. But in a really controlled environment, like when they call in and it's got like very clear parameters, it's pretty dang safe.
Fer Patel: Yeah, those who don't know much about AI or the beginners into it, may the concept that a lot of people will come come to experience in a unfortunate sometimes is when it hallucinates.
Danchez: Yeah.
Fer Patel: It starts making up things all of a sudden.
Danchez: Yeah.
Fer Patel: Like, where did that come from?
Danchez: yeah
Fer Patel: That doesn't make sense. The hallucination all depends on what model you're running under, how much contextual window you have, what do you have in your knowledge base, because your AI is always trying to answer whatever question comes its way.
Fer Patel: Right or wrong, it's going to answer it. So this is where you got to put all the guardrails in place saying, if you do not have an answer for this question, do not make it up.
Danchez: Yeah. Yep. Yep.
Fer Patel: Just say, I'm sorry, i don't have that information. I'll get back to you. Don't try to make it up.
Danchez: yeah
Fer Patel: So this is where proactively we would like the agents to inform you, tell you things differently. you know if it If you sent an email the month previous and you generated X number of dollars in revenue saying you sent an email last month, it did you made $25,000 in revenue, here's another email I've generated, similar idea that you can send to this smart list.
Fer Patel: It's going to look at your metadata and give you suggestions versus you having to say, I need to send an email to this list. This is where we want AI to go to become helpful to you to make better decisions.
Danchez: It's funny what you just mentioned is something Sam Altman has been saying out on the road as he's done interviews. Essentially, like AI will just be proactively telling you, hey, it did this and this. Do you want it to do X, Y, and Z next?
Fer Patel: Yeah.
Danchez: That's just kind of the future of where I think a lot of people are seeing AI going now. And I think we'll all have like a co-pilot with like ChatGPT kind of helping us through our personal tasks. It'll be working us through so we can get through faster.
Danchez: But I can almost see like everybody's going to have like a agent that's attached to their CRM because it has to be. Like your your business revolves around... the relationship management system, the database, the information, the source of truth.
Danchez: Therefore, your CRM will be a main component. And I'm like advocating that high level should be the one for some people it's going to be HubSpot. hubs And HubSpot's been making some significant substantial gains with theirs. I just find that it costs too much.
Danchez: um
Fer Patel: No problem.
Danchez: I'm so excited about where it's at now, but where where it's going in the future.
Fer Patel: Awesome. I'm excited as well. It's going to move very fast.
Danchez: And one last time, if you're still listening, it's 615-610-0721. Call it. You will be amazed at what it can do. Again, not perfect, caveat, but you' I think you'll be impressed with how far it can go.
Danchez: Ask your questions about high level. Book a meeting with me. I'd love to talk to you about how i'm using it. And just test it out. Take it for a drive.
