Jimmy Huber: And welcome to another episode of the Info Tech Podcast, where we explore the intersection of business and technology. I'm your host, Jimmy Huber, and I'm sure you guys are familiar with the ever changing world of AI. And we're gonna revisit that topic today with my guest, Amir Pivandi, with Aventura. Amir, thanks for coming on. I appreciate your time.

Amir: My pleasure. Nice to be here.

Jimmy Huber: You bet. So we were chatting a little bit before, and if you're a regular listener, we've talked about AI before, but not in the space that Amir works in. Amir is in the heavily regulated area of banks and and all these places where security is paramount, right? so we've got a lot to talk about today. Amir, but before we do that, why don't you just give us a little bit about your background, maybe how you got started in business and and what made you get into the AI space?

Amir: Absolutely. Thank you again for having me here. And really if you go back thirty years ago when I started in insurance domain in Canada, we've always known that insurance is highly regulated but at the same time underserved. It's one of those platforms that due to just the nature of the business they're conservative, they're adopting technology not as fast as other sectors. And three years ago or so, you know, being AI all all over the places, I came up with an idea of what To do to make sure that insurance can adopt AI in a faster pace. And I started Aventura, built a bunch of platforms and been trying to make sure that we can in this industry, which is highly regulated, we can take advantage of AI without compromising security and personal information and making insurance leaders comfortable adopting AI technology.

Jimmy Huber: Yeah, I think that's great because without that, I you seem to create more problems than you fix, right? and you've gotta be on the bleeding edge because you're the first one that I've heard that's in this space and already doing agentic. Usually people are still in the, you know, the generative world, you know, the chat GPTs, the claw. but you're already getting into the agentic part. So describe that for someone maybe who's not familiar with with the agentic model. I mean, you hear all the buzzwords and everything, but What does that mean if you if one of your clients is working with you, practically, how does that change, you know, their world?

Amir: I think the Best way to describe this is to explain how basically the platform right now works. We have two main platforms: one is Engagement Hub and the other one is Client Grow Growth Hub. And both these platforms use Agentic AI in a way that we build and define a target for AI, collecting the data but using the data, making sure that there is nothing in it that identify people, identify entities. put anybody at risk, but then give enough to AI to be able to conclude and come up to some Results in terms of how to adjust the behavior of the application and the way it works in order to get better results. And I give you an example, right? So one of the things that our system does is when people go to aggregator websites or organically an insurance company or a mortgage brokerage, they find they detect the signal, put it this way, that someone's interested in their product. Our engine kicks. In we start communicating with that per that person, that entity, and of course, we're using generative AI in order to you know have that conversation. But the part that is important is detecting the signals based on the tone of the conversation that is happening, based on the keywords that is being used, based on the profile that we're building, and this is not the customer profile because we're trying to not even depend on it for security reasons. For privacy reasons, but more importantly, ability for the system to think for itself. So it gets and detects all these signals and adjusts its behavior. It it may, you know, decide that this person at this point in time is not ready for that transaction. And simply instead of pushing for a sell, say, Yeah, I understand that you know you're not ready at this point in time. Is when is a good time to reconnect? Is it next month? Is it three months from now? This is just Just a simple example how that there is a difference between just having a generative AI, building a conversation, and pushing and pushing and pushing and going forward, like someone like a salesperson that calls you, versus having someone that has the intelligence beyond the scene to understand your emotion, understand the timing, understand the conversation, get those signals, but more importantly, use this to adjust its behavior going. forward so not only the system adjusts the behavior on the call on the conversation that is happening at that point in time it also uses this data for later conversations with other people based on the trend that is happening it basically collects that data and it applies to its behavior and that's basically what what what the system does.

Jimmy Huber: That's amazing. So you're basically taking inbound calls or doing outbound like cold calls, or where does that fit in the sales cycle then?

Amir: We do both. So we started with doing outbot calls. So the very first client that we had, a a large brokerage in Toronto, they are purchasing leads today they're still a client. They're purchasing leads. And they had a small team that they needed to basically reach out to these leads and make sure that they're actually real people. They actually need, you know, need a product. And in insurance domain, 10% is basically the sweet spot of the leads that you get, become a customer. But in in order to get that 10%, you have to engage 100% of the leads. So we started building this platform To go and engage these leads and only connect to the call center people who are actually willing to talk to the call center. So tons of the calls, tons of, and we start normally with SMS, then we switch to call, then we go with SMS. So tons of these calls go cold, but around 17% at this point in time, the person either says, Call me tomorrow, I'm driving my kids, or it says, Yeah, I'm ready to talk to the call center. Then we connect them to a licensed insurance broker to talk to them. So this was the very first product, and then we added the inbound cult. And now the inbound cult is gonna be you calling to the call center and you can ask questions. And these are general questions, again, highly regulated environment. We we are not supposed to give pricing, we're not supposed to offer anything in Ontario and most of the North America. You have to be licensed in order to To be able to have that conversation. So this is this is what the system does to make sure that you know it doesn't cross those boundaries. And the last item that we've added recently, about three months ago, and we're basically expanding it right now, is moving away from that lead chasing, lead nurturing platform, which is engagement hub, to lead generation. And now what we do, we basically start with social engagements, we use Reddit, we use LinkedIn. We use Facebook, we use Google, and also use Perplexity's long research platform in order to find

Jimmy Huber: Yeah.

Amir: businesses or consumers.

Jimmy Huber: Okay, so that was a ton right there. So basically if if I would boil this down, you are helping frontline salespeople be more efficient in their initial intake and their outreach, right? So you're you're probably

Amir: Absolutely.

Jimmy Huber: answering a bunch of common questions and getting people into a a funnel to then hand off to the human when they're in a position the timing is right. So you're probably saving them what, sixty, seventy percent of their time? 'Cause they don't have to chase people down. They're only talking to people that have already been vetted out and are ready to have a conversation. Is that right?

Amir: Absolutely. But even more important, think about this, right? The one of the factors that technology people may be not paying attention is the human factor here. Right now we have call centered people that they were miserable, they had to do cold calls, they had to talk to people that they didn't want to talk to them. What happened? We turned that group to work with people that actually want to talk to them. So they don't reach out to anyone. that doesn't want if they talk to someone

Jimmy Huber: Yeah, that's huge.

Amir: this person wanted to talk to them so the happiness factor is higher but more importantly we're giving a lot more to the customers because not everybody really wants to buy from you but everybody needs your expertise needs your knowledge have questions they want to get some sense and you don't need to deal with them we deal with them you only focus and put your efforts on actual customers that are going to create ROI for you.

Jimmy Huber: Yeah, I love that. I mean, I've done cold calling even for infotech with our RMSP and it's a mental challenge. you

Amir: It is.

Jimmy Huber: know, I mean it's a numbers game and that it wears on you over time. So I can totally see the value, you know, for your customers there. What about the the the inbound people, right? So there's a lot more of these AI features, right? That you you pick up a phone and it takes a while to get to a human sometimes. And I think that's probably not completely normalized yet. So what you say to the people that are like, hey, like I don't want I I want my people to talk to humans off the get go, right? Like I don't want them to have to to go through an AI thing that people think it's human and then maybe it's not, right? There's there's this I think negative connotation sometimes when they're not taking into effect what you've said, like picking up on the social cues and the emotions and all that. I mean, they're probably thinking of voice agents from, you know, a year, year and a half ago. So what are you seeing with people and how are they responding?

Amir: Yeah, that's a that's an excellent question. And you're pointing to a very important factor here, right? First of all, it's wrong to get a technology and start adopting it everywhere, right? There's a there's a place, there's a use case for this technology, and there are places that it is not, right? So imagine that you're calling your doctor for an emergency. You don't want to talk to any AI. You really need to talk to someone immediately to give you attention, right? But

Jimmy Huber: Sure. Yeah, of course.

Amir: if you're calling to ask about your insurance and find out you know what can you do to lower your insurance costs what are the options or if you're calling to you know your travel agents and you want to know about the new destinations or products if you're you know doing other stuff that you are willing to do the research you actually don't want to talk to people right so you're you don't want to talk to someone to who wants to sell some because you're collecting information at this point in time. So again,

Jimmy Huber: Got it. Yeah.

Amir: this is this this has its own use case and place. And by the way, depending on the you know what age group you are, some people and some, you know, generations are much more comfortable with AI than than than than humans. So this is not a you know done and used for all solution. This has to be something that it's required and is needed and you always have to give the option to the customer to talk to some someone. And this is you see it more and more and more AI companies are adopting this that allowing people to request to talk to an agent. And even then you have to collect some basic information that you could do and answer some basic and you know questions.

Jimmy Huber: Yeah. No, I love that. The use case is is definitely key. And I I think people think it's a one size fits all. And then they kind of fall flat and then they assume that everything is like that. They kinda lump in everything together and and it puts a bad taste in their mouth. But yeah, I think educating people on the best way to use the specific, you know, software in the case that it's designed for is is is super key. You mentioned privacy

Amir: That means

Jimmy Huber: before too. I wanna get into that because you know, security around AI these days is is just bonkers, right? People I don't think they realize what they're doing when they're using a a chat, even a paid version that doesn't have the correct security around it. so you said a few things earlier about, you know, picking up on the you don't get people's personal information, right? Especially in these industries like healthcare or whatnot that have, you know, PII is a big deal, right? so how do you train the the AI to get better, right, and to learn and to use the data while fencing off those privacy needs? That's gotta really be difficult to thread that needle.

Amir: Absolutely. And we we we had this in mind from the day one. And again, because I'm coming from you know that highly regulated industry. And one of the things that we've done is there is t two two aspects, right? There are two misunderstanding when it gets to the AI, and I wanna make sure that we're clear on that one. First of all, you don't need, you know, general AI to go and you fix a very specific problem. So For example, if you're building an agent that its job is to answer insurance questions about you know Ontario auto insurance and help you to connect to a sales agent. This AI doesn't need to have a lot of knowledge that requires you to go and use open AI or Anthropic or other Gemini or anything else. You could very well build the solution using a low car LLM running on a local data center. So that's one of the things we do. If you go to Aventura website, we have a concept we call it AI in a box, right? So from

Jimmy Huber: Yeah.

Amir: day one We built the system to be agnostic to the LLM. We built the system to have the ability to run on a local network. And of course, you know, you're gonna get a better performance if you go with a larger model and more expensive hardware, but this doesn't limit you to protect your data and privacy of your clients if needed. So that's that's number one. Number two, which is important, is we built a technology that filters sits basically between the customer data and the LLM. And what we do, we use this technology to remove all the personal identifiable information from the system. So for example, just give you a very simple example. I want to build a prompt to greet you and let you know that you know we're ready to talk to you. Do I need to really pass your name to this, you know, LLM? No. I can I can say ask customer. This and get that, and then inside our secure environment, we replace your customer, the customer keyword with your name. And this is just a simple example. We do this with their date of birth, we do this with any information that we don't want to pass to LLM. So it's not that the LLM is secure, we don't even pass the data to LLM to start with to be worried about security. So either you use the AI in a box on a local environment. Environment on a limited knowledge and you train it and everything works, or you go and you make sure that you use a smart design not to pass the information to the LRM. The only caveat here, just to be honest, because I'm sure your audience is gonna say, what about this? And they're right. The only caveat here is when you use the voice agents, and this is The part that you turn text to a speech or speech to text. And even for those, we're working with DeepGram and we're working with Bedrock at this point in time, AWS, and they're coming, they're doing their part as well. We're we're a software company that uses AI technology. We don't build AI technology, but we collaborate with companies and big giants that they do build the AI technology and they get our feedback and we're talking to them yesterday I had a meeting with AWS exactly on the voice agents and that was the topic. How can we make sure that we have the least latency but at the same time we don't pass that data that we don't want to pass.

Jimmy Huber: Yeah, that's super clever. I think that there's a lot of value in doing what you're doing because if you're building the AI yourself, there's a lot of overhead and I think there's lots of people already doing that, right? I mean, it'll be interesting to see how it shakes out and how many companies survive, right? But the value is being able to connect the dots for business owners and to build things with the existing platforms that are out there at like exactly what you're doing. So yeah, I I give you guys a lot of credit because There's not a whole lot of people that at least I've run into yet that are taking it this far, right? There's almost like a race just to get out there and to be in the game, but you're already looking around at what's available and saying, all right, how can we optimize this and put some things together to address a a niche that, yeah, has a lot of security concerns and yet has a bunch of needs that, yeah, I mean, I love the AI cold calling concept because there's certain things that are humans are really good at. And then there's certain things that really beat us up that AI can just handle, right? So it's all about the use case and and maximizing it, you know, for your customers. So I I absolutely love that. yeah.

Amir: That's that's the keyword that's the keyword use case. You just use it where you need to use it versus just apply it to everything just because you wanna have AI.

Jimmy Huber: Yeah, I love that. So fast forward a little bit then. I mean it sounds like you guys are just off to the races here, but you know, it's changing so rapidly. What are you guys working on, you know, for the next six plus months? I mean, do you see this optimizing or getting into other industries or where do you see this going?

Amir: We're always open to new industries, we're having conversations. one of the areas that we're hoping to get closer to is healthcare. we do have support for clinics and dental offices and everything else, but I think healthcare by itself has extra challenges when it gets to privacy and when it gets to securing the data. So that's one of the areas. But for the next six months, Aventura's plan is to focus on our client growth. That's the lead generation platform because one of the things that we found out is a lot of businesses are thirsty to get the right lead and prospect to start that communication. And at this point in time is a crowded you know environment with a lot of players excuse me but none of them are using AI the way they should use it. So we build our client growth hub based on that ideology that we need to be able to go and r find the right people. I I I I tell you one thing that happened last week which was interesting. So we have these agents that they go so we define basically the ideal customer profile ICP and this agents go and start Doing the running, you know, searches on, as I mentioned, perplexity, deep research engine, and Reddit, and so on and so forth. When we detect the signal, we start communicating with these people. We start sending email and then asking for their phone number and so on and so forth. But the funny thing is that when for the third communication was a phone call that the system made, and we're looking at the transcript, and the person, after having a conversation, at the end of the conversation, the person says, can I ask a question? Yeah, sure. Are you an AI agent? And that was like I was looking at that and I said, this is exactly what I want to see. So not only this was done in the right way that the person agreed to have a conversation, but they didn't even detect, but they kind of had a sense of they're ta not talking to a human. And that's the type of basically environment that we want to get to, and that's going to be our focus to make sure we get more and more of this type of experiences that people feel comfortable talking to AI. Deliver business value, and I think it's going to be a win-win-win situation for the businesses, for us, and for the consumers.

Jimmy Huber: Yeah, I I think the cycle to normalize that is coming in a hurry. way faster than than other technologies, you know, in the past. And so I would I'd be interested. So when they said yes, I'm an AI agent, what did what did the people say? Were they like, this is great and kept going? Or 'cause I

Amir: Yeah, no, the it it can it was at the end of the conversation and it it basically

Jimmy Huber: Okay.

Amir: was a confirmation of what they said and the the trained agent basically confirms so yes I'm an AI agent, but if you want I can connect you to a person right now and that's the key. The key is to give option and I hate when you go to the some of these systems and you have to there is absolutely no way to get a to a person. This is not to replace people, this is to make it simple. simpler for both sides, for the call center people to talk to happier people, spend the time actually providing value, and for the consumers to be able to get to the answers much faster, much easier. And again, everybody's gonna win in this.

Jimmy Huber: Yeah. I I will say it in my own world, just to get guests and find people like you for the podcast, I've used some AI tools, you know, for our email funnels and that. And it was interesting because we haven't done a ton of that at Infotech, right? Like we're still doing a lot of human outreach and, you know, referrals and that sort of thing, the traditional methods. But I would say it's less than one percent have ever pushed back when sometimes it's almost obvious that it's AI, but people still just keep going and it's like, wow. That surprised me. I thought we'd have some problems, you know, especially in the earlier days when before they were refined, but I I haven't seen hardly any resistance. So it's almost kind of scary that people are just willing to deal with, you know, software basically, at least in some areas, until they get to a human. So I agree with you. I I think there's a best of both worlds and there's a way to do it in the right way, right? Where you give people the options and then you can kind of have a win win scenario. Would you agree?

Amir: can't can be in agreement more and it's it's it's one of those areas that first of all I don't think there is any way to avoid it. This is gonna happen.

Jimmy Huber: Hm, yeah. Yeah.

Amir: This is like internet. You know what I mean? This is gonna

Jimmy Huber: Yeah.

Amir: happen and we can do it right, and we can we can basically get to the right side of this equation or not. I I know there are concerns, I know there are worries, and those are those are valid, and those are things that we have to be concerned and we have to be worried. But that said again, I repeat myself: use case. Find the right use case and use it. Don't give control. of your entire PC to AI, but you do get, you know, give control of very specific part of your work that can make it easier for you to achieve the goal that you have versus like, you know, spending time doing things that can be done much easier.

Jimmy Huber: Yeah, no, I love that. So a few minutes left here, Amir. what would you say to maybe the younger entrepreneurs out there that are are starting to get into the AI space and they kind of don't know where to turn, right? It's such a a massive opportunity, I think. you know, if you're wanting to be a developer or getting into those traditional things, you better have AI skills or you're not gonna be near as valuable in the marketplace, right? So you and your team have been doing this for a while. Obviously, you've had some success here and and are figuring things out. what's your best advice to those that are maybe just a little bit further behind?

Amir: I think the advice is the same for AI or not for AI, for technologies people, right? Don't build a solution and then find a problem for it. Find a problem and build a solution for it. Find

Jimmy Huber: That's great. Yeah.

Amir: a business issue challenge. And try to solve that. And if you can use AI to solve it securely and the right way, perfect. if you can use AI to build the solution, not including AI and wipe coding and these days building application, it can't be, you know. simpler and easier these days to go ahead and get an idea and start building it. It's before, you know, it took time. Right now it's much easier. So maybe that's the angle to look. But again, the most important thing is to look into the value that you bring in. What is that problem that you're trying to solve versus doing technology for the sake of technology.

Jimmy Huber: Yeah, I love that. I think their shiny object syndrome is running rampant these days. I mean, even at the larger companies, like the iterations of putting out new versions of models seems to be almost weekly, if not daily. Sometimes you're always seeing new stuff come up. And so, yeah, doing things because there's a need and value and you're fixing something versus this is this is cool. Cause that that's where I tend to go. It's like, wow, I didn't know I could do this. Well, did I even need to do that? Like, no. So Having some thought and and making AI, I think, a thought partner and not necessarily a thought leader, right? Like you gotta have a little bit of both.

Amir: Absolutely. Absolutely.

Jimmy Huber: well, thank you so much, Amir, for being on. I I really think that our audience had a ton of wisdom here to glean. if they want to get a hold of you, whether it's you know, to talk about the the product, the aventura, or to just, you know, get in touch about AI in general and start a conversation, what's the best way to do that?

Amir: website if you go to our website aventura.ai there are tons of information also if you want to learn on the technology side we have a docs.aventura.ai that we basically openly share tons of information about how we build this application how to use these applications and my phone number and email address is there and I'll be more than happy to talk to your audience.

Jimmy Huber: Be fantastic. Yeah, we'll we'll definitely put your website and and link in the in the show notes here so that they can find you easily. and yeah, I mean this is this has been great for me. going through, you know, this is the we're winding down here the first season of our podcast, but I'm definitely gonna add you to the list of people to talk to in six or twelve months because there's just so much changing that we want to be able to follow along and and to keep up, you know, with with not only your story, but you know, the the AI scene in general. I think people that are Embracing it and not pushing back are gonna be only more benefited going forward and and be leaders, you know, in the industry. So thanks for thanks for sharing.

Amir: Absolutely. It was a it was a pleasure and looking forward to further conversations. All the best. Take

Jimmy Huber: You bet. Take care.

Amir: care. Bye.