Jake Aaron Villarreal: I'm Jake Aaron Villarreal, born and raised in Silicon Valley, and here to take you behind the scenes to share what it's like to be a startup founder, the journey they're on, the products they're building, the problems they're solving in an effort to transform industries. I'm excited to have with us today Suman. Welcome to the show, Jake.
Suman Kanuganti: Glad to be here.
Jake Aaron Villarreal: I'm glad to have you here. We've had a lot of conversations with a lot of companies in the AI space and they're building and they're growing and there's a lot to to learn as we go. Uh before we jump in and talk a little bit about your company here, a little bit more about uh Suman. He drives Personal AI's vision forward, transforming how organizations adopt AI. Prior to Personal AI, he founded and scaled Aira, an AI/AR accessibility technology for the blind. Suman was named in Forbes 40 Under 40 and is renowned, a renowned leader in accessibility technology. You've got a great background. I love the focus you've had and and curious to learn a little bit more about the decisions that you've made to really build and and and and fund or found Personal AI. Before we jump in here, where are you calling in from today?
Suman Kanuganti: I'm calling in from San Diego.
Jake Aaron Villarreal: Hundreds of AI startups are launching every month, battling to build their founding teams. As a leader, your job is to get results. When it comes to hiring, that's where it gets tough. So, you go out and you try a recruitment firm, but they don't understand your story. They're off target. And when they send you candidates, it's a waste of time. We believe you should never have your time wasted. That's why we launch Match Relevant, because your story is more than just an open role. It's your founder's journey, the problem you're solving, the product you're building, and why it matters. When we work with companies, we make sure we understand your whole story. So, we go out and do a search, we're on target, it's worth their time, they're interested, and more importantly, it's worth yours. And when it comes to hiring engineers, we work to make sure we get it right by deploying a team of seasoned CTOs that have built some of Silicon Valley's best companies. They can collaborate with you in the technical interviewing process. They can be a sounding board or they can run it for you. When it comes to building teams, there's no time to waste. Let's make it count. If you have a role that needs to be filled, book a time with a hiring guide at matchrelevant.com and learn how we do it.
Great. Love San Diego. Went to school down there. It's a great place to be and live and innovate.
Suman Kanuganti: Which school did you go to?
Jake Aaron Villarreal: I went to San Diego State. Was, it was the party school. Not the really smart school that you're at today.
Suman Kanuganti: I went to UCSD, rather, but actually in UCSD campus right now. We have an office in the campus.
Jake Aaron Villarreal: Oh, very cool. Yeah, I I uh I loved San Diego State. It was such a great experience. And ironically enough, um it was a very social school and a lot of fun. Um it's really hard to get into now. So, look, makes me look better. Um, because it was a little bit easier when I went to school there. But, there was a lot of founders that were in my network when I went to school down there. I was happened to be in a fraternity, but you know, the the CEO and founder of Rockstar Energy Drink was there. He was a fraternity brother of mine, sold out for $4 billion. And um, you know, the founders of Volcom, one of them who was the design lead there was a fraternity bro. And it was just a lot of um social innovators and looking back now um it was, it was really fun to be there. I'm now in Orange County, so not too far away from where you're at, but...
Suman Kanuganti: Well if it is helpful uh my co-founder Kristie Kaiser, she's from SDSU and I have two more people um from SDSU as well, so you know we are all friends.
Jake Aaron Villarreal: Yeah it's great, there's, there's such a... it's a university town in San Diego. There's, you know, a number of universities and it's a young town. It's a lot of, a lot of smart people making uh good products down there. So, really excited to learn more about your company here. And before we do that, give us a little background. We'll dive into your origin story a little bit. Um, you know, how did, how did you get into technology?
Suman Kanuganti: We are going to the origin story. Yeah. Well, you mentioned San Diego as a university town, right? So the town that I grew up back in India was also a university town. Uh it was like 2 hours away from the metropolitan city called Hyderabad in the south part of India.
Jake Aaron Villarreal: Oh yeah.
Suman Kanuganti: And we only knew one thing when I was growing up. You either study or you die. That's at least what my parents taught.
Jake Aaron Villarreal: Yeah.
Suman Kanuganti: I mean not that I was forced into studying a lot, but I uh definitely enjoyed doing so. Um and part of uh me driving through engineering school in India uh brought me to the United States for continuing my education in robotics. So I was an electrical engineering background major and then I came to University of Missouri-Columbia for getting my master's in robotics. So I do have a pretty intense engineering background, if you will, all the way from uh low-level systems to systems engineering to now you know software as well as network level engineering as well.
Jake Aaron Villarreal: Yeah. Yeah. You know going back to your background a little bit, what was the first job you ever had? Was that in India or was that here in the US?
Suman Kanuganti: No it was here in the US. Uh that was immediately after I finished my master's. I went to Caterpillar.
Jake Aaron Villarreal: Right.
Suman Kanuganti: Caterpillar. If you think about an engineer's dream job, was down on the manufacturing plant programming robots. Um, I had such a fun time. Dude, this was back like, you know, 20 years ago, right? And for the fact that you are uh playing with these giant robots to move these heavy duty engines from point A to point B and you're able to program them, it's like a dream come true, if you will.
Jake Aaron Villarreal: You know, it's funny you say that. Um, you know, John Deere is a company that, um, we hear a lot about when we talk to robotics companies that they're looking for talent, you know, and they're going back to these large organizations that you wouldn't necessarily think, you know, have been innovating in the AI space or in the robotics space or in general the data space. And they have done a lot of investment, and really a lot of top engineers are coming out of companies like that, especially now with AI being so prevalent of being able to adapt it to different areas um within different industries. So uh it's, agriculture seems to really be an area where there's a lot of great innovation and um...
Suman Kanuganti: I always um, I always think the change of adoption as well as like rate of adoption to new technology is almost like directly proportional to the amount of assets and operations that you have to drive to keep your current business going. You know what I mean? Uh one of the reasons why like John Deere and Caterpillar, they have to deal with tons of existing assets such as your manufacturing plants, your robots, your processes and they have deliveries to make right, and then on top of that you have to adapt to the new technology which becomes challenging because then you're also not, you know, you're also dealing with like the physical assets, not just you know digital assets like Google and Facebook if you will, right? So yeah, so I think it's not surprising. I think uh as the ecosystems kind of evolve into uh startups and uh incumbents can adapt or embrace or acquire the startups to next generation of technology, I think is uh kind of what you know the ecosystem is made for.
Jake Aaron Villarreal: Yeah. Well, let's talk a little bit about the ecosystem. You're playing in a space where there's a lot of innovation in AI, and um you know usually we look at the market and say, "Okay, what problem are you solving and then are you building something that's worth paying for?" So talk a little bit about Personal AI and what was the inspiration behind identifying what problem you wanted to solve and what are you doing in that space today?
Suman Kanuganti: Yeah, let's let's talk about the problem in itself, right? Uh and this problem that I've been solving which I will talk about isn't new with Personal AI for me. It's been in existence for the past 15 years of my professional career, which what I refer to as the human capital problem. I was at TurboTax for 6 years, right? If you think about accountants and their jobs and their roles and the demand that they have year-over-year business to be able to complete and fulfill you know 200 million citizen taxes, it's a big undertaking, right? So the way Intuit and TurboTax has driven the business is how can you drive software economics to a human capital problem, which is the accountants, right? And clearly, you know, it's a big winner now, and it you know creates like tons of value to scaling those um accountants. So I always kind of emphasized and picked up that core idea of bringing software economics to human capital, starting off with accountants at TurboTax.
And then my previous company you mentioned Aira, which is an AI company, but AI was in the back end. Human agents were in the front end for people who are blind and low vision to provide remote assistance by teleporting into their, you know, physical environment, physical world. Right? So if you think about what the problem we are solving over there, is very similar, which is a human capital problem where you need a sighted assistant next to a blind person to go about things. Either it'd be working at a company, either it'd be simply walking to Starbucks and grabbing a cup of coffee, or like flying around in the country um you know, going to places. So in that scenario, we basically replace the need of having sighted assistants but provide Aira uh as a service to scale or bring the software economics to that need. That's also a human capital problem. Uh today uh pretty much every big tech including the likes of Microsoft and Amazon and Comcast, uh every airport in the country, all the 100 airports in the country, plus every uh bank such as Bank of America, every Starbucks uh in the country are Aira accessible locations, right? So suddenly you have opened up the need of human capital to a much broader you know, need and experiences to unlock.
Personal AI kind of pushes that idea of bringing software economics to human capital to the extreme, to the idea that how can you create a model to unlock the applied knowledge and the experience of an individual person that they grow up with. Right? This company started in 2020 with a core premise of being able to bring the software economics to yourself, to your own applied knowledge. So that way the people around you, the team members around you, and for your own self, you are working with your own persona, we refer to them. That is the applied or application of your own mind to the newer things that you are working and doing. So the problem that I always solve for is the human capital problem. How do we scale human beings to do more?
Jake Aaron Villarreal: Yeah. So you're basically saying how do I scale myself? How do I make more of me? Being able to do a job and having technology do that for me based on the knowledge that I know that it can then communicate to whoever is asking me questions about my business or about problems I need help support with or whatever it happens to be. Is it really almost like a digital twin of me communicating with the work that I have to get done that's sitting on my desk?
Suman Kanuganti: At the simplistic level, yes. Right. But you can push it a little bit farther, which is: yes, sure, I can ask Jake a question and get a response back, right? But then I can also ask Jake um a new concept or a new context and how would Jake think, how would Jake synthesize, what would be the potential decisions that you would drive towards and for what reasons, right? So we do get carried away from question and answers from a customer support standpoint, but what we see the application of Personal AI that is happening in the real world is the depth of applied knowledge that an individual role or even a functional role, it doesn't necessarily have to apply to an individual, but in a small business setting or a medium-size business setting, you could create a marketing officer who may not be hired or recruited as of yet. But generally, the founder that is kind of driving the principles on what that marketing persona would need to be, you know, to work with that marketing persona in advance of even recruiting somebody.
Jake Aaron Villarreal: Yeah. Really cool. You know, Reid Hoffman, the founder of LinkedIn as well as now, you know, he's been in the VC world for a long time, you can see these interviews that he does where he's talking to his self 20 years ago and his, you know, persona of his younger self is responding based on hundreds of hours of video training on how he would think and how he would respond. Uh, it's really engaging. It's also really valuable. Um, I think of that as, God, if you could have that, that's really encompassing how you really think and a lot of data that's trained on how you think, but then applying that to solving problems that or doing work that you would have to do day-to-day. So, um, when I when I think of the concept you're talking about, is part of that... is like any of that kind of in there? I'm just trying to make sure I'm getting it.
Suman Kanuganti: No, that's 100%.
Jake Aaron Villarreal: Okay. It's incredible. And God, if you could create hundreds of those, I mean, you're kind of hiring bots or agents instead of hiring people and scaling yourself, but also scaling a company or an industry really.
Suman Kanuganti: Maybe that's uh, that's, that's like literally what we do with our companies, right? If you think about, you know, who are the next five to 10 recruits that you are potentially going to hire. And the easy way to think about Personal AI and the persona is no different than hiring somebody, right? What do you do with hiring? Okay, you have a role, you have a definition, um, and you go out and you recruit the person. Once the person comes in, you train them, right? By training them with existing knowledge, and then you have like specific workflows and specific tasks you associate with it. But it also is dependent on the applied knowledge of that particular role. Right? So the existing, you know, employees then suddenly does you know, 5x or even 10x sometimes more amounts of workload, you know, either it be cases for an attorney or either it be decisions for a corporate strategy person, right? They are doing five to 10 times more um than otherwise what they would. So they become technically, like, persona managers and kind of embrace that mindset.
Jake Aaron Villarreal: Yeah, that's really cool. Is the technology here now? Is it in, is it in production? Are companies using it? Are they learning from it? Where is the, where is the company at now? You started 2020, correct?
Suman Kanuganti: Yep. We have tons of companies who deployed Personal AI into their companies with uh tens if not hundreds of personas that are doing real work and real job. Uh we have evolved into companies hiring or training persona managers from a human uh labor standpoint, you know, somebody needs to still manage, right? Uh so it's quite fascinating. And uh yes as you said, the company uh inception happened in 2020 and since then we've been uh evolving. So we've been through the pre-AI era and post-AI era, uh so it's quite fascinating you know personally for me like being inside the industry and kind of you know being part of the change if you will.
Jake Aaron Villarreal: Yeah. Yeah. You know, when you talk about being kind of in the development stage before, you know, really the world heard about generative AI and OpenAI and these companies that have been building, and now being adopted regularly across lots of industries. Um, we kind of look at it as Lego blocks that are able to be built and produce services or products that a lot of companies can get access to now. And so the question then becomes, are you building the Legos that are provided from the big companies that you can carve out your own niche or are you building from the ground up your own technology stack? Walk us through how you're approaching it.
Suman Kanuganti: For us, it's the latter because this is a company where we wanted to offer the Lego blocks for organizations and for individuals to come up with their own um creative sets of tool sets, if you will, right? Um you are right, like OpenAI kind of provides like building blocks, Personal AI also provides building blocks, right? Um now there is an approach uh and the difference that we kind of took which is Small Language Model from a technological standpoint instead of Large Language Models. We heavily emphasize or focus on personalization use cases, Personal AI, right? Um is this core idea of like unlocking the applied knowledge and applied thinking, how a subject matter expert or an individual person would do, with that vision of like everybody would have their own personal AI, right? And for what it's worth, "personal" is a category now. It's not just a company.
Jake Aaron Villarreal: Is it? Okay.
Suman Kanuganti: I mean, yeah, personal... pretty much every company has a personal AI that is uh going on and I can, we can talk more about that, but to uh answer your question, you know, we... back to the uh analogy of like Lego blocks, we looked at, oh, Large Language Model provides some of these Lego blocks as well, right? So then we... then our approach in 2022, 2023 is kind of like, let's embrace the newer Lego blocks because, you know, people are building on those Lego blocks, and bring those two sets of uh foundational Lego blocks together, right? Because people need personalization but people also need to understand what general knowledge has to offer outside their own realm, outside their own organization. And the beauty now in here is I think the timing is perfect where Personal AI, the personalization engines has gotten so mature, and uh the LLMs gotten so mature, the goodness of Small Language Models or, you know, we refer to them often as Personal Small Language Models or Personal Language Models on the branding side, the goodness of those two highly mature, you know, looking inwards and seeing through the world through the lens of how you would think, um is quite fascinating. So even like, even in the last uh 12 months, these two worlds coming together, the rate of use cases and the quality uh and the change that I see in the industry is quite fascinating. Uh so yeah, to close the thought, now it's almost like you have a, you know, two brands of Legos that fit together and you can unlock even more things for yourself.
Jake Aaron Villarreal: Yeah. To build on. Yeah. Yeah. You know for the audience that's hearing this for the first time, you know, we're going to have a a sample of what this actually looks like and sounds like and interacts, and we'll add that into the episode. But, you know, what are the what are the sectors that you're focused on? You talked about legal. I know you were talking a little bit about financial as well. Um, there's a lot of companies, a lot of founders that are listening to this like, they're probably thinking, "Hmm, how can I have a, how can I scale myself or my team and use Personal AI?" Are you dialed into a specific sector or is it you know, "Hey, any company that has a role we can we can train a persona for that company"? Like what's, what's your your focus?
Suman Kanuganti: Um, our focus is in the compliance markets. What do I mean by that? Persona is really good to unlock the applied knowledge that otherwise is not available in the public corpus, right? So when you think about compliance markets where people have ultra-sensitive information. Uh people cannot, you know, technically give that information to anybody even if you are a partner, right? Those are the areas that Personal AI is really good at because not only we train uh the models on their proprietary data using this persona architecture that I was talking about, but also we give the ownership of the data and the model to the company, to the individual, or to the organization.
The reason is we, from 2020, had one core promise of Personal AI. This core idea of like democratization is not necessarily accessibility to the technology, but accessibility to the technology that they feel comfortable, safe, and secure and they get to own it as an asset, right? Like uh when you buy things at home like your TV, you buy it, you keep it at home, you can break it, you know, you can do however, how use however you want. Uh so, so we we give this uh model ownership and the data ownership and there is zero chances for the memory to be used for training any large language model, right? We scale horizontally. Every organization gets their own model. Every persona within an organization member is siloed from each other. So there is no cross-contamination of the data, thus reducing hallucinations as well. But more importantly, increasing the accuracy, but the accuracy from a sense of uh context that surrounds a role, could be individual role or a functional role.
Um I can give you an example. Let's say you have an attorney in a medium-size law firm that generally deals with lots and lots of cases. But most of this attorney has very um, they they have specialization in specific, you know, case types. It could be you know, drug-related cases or murder-related cases or, you know, other things. And and there is a historical context associated with how they deal with a new case that is coming to them. Right? So when you think about a persona of an attorney who has specific types of cases that they have built, have um figured out their own strategies, the secret sauce of their own companies, and how they write their arguments, how they unlock the evidences, right? Uh so when you have your persona to be able to deal with newer cases that are coming at you, either it be you know, drafting the arguments based on all the historical cases that they have done or unlocking the insights on, you know, potentially um you know simply even listing down what are the, what is the pieces of evidence that I need to, you know, gather based on the past winning cases. They become highly specialized, highly complex that has in-depth rooted knowledge, you know, into the firm's knowledge or into an individual's knowledge, right?
Um, so one of the biggest differences between what we solve for and what majority of the large language model solves for is we deal with high depth context uh and specificity that is associated with a specific workflow. It's almost like you know, tiny vertical-focused models that are many, many, many in millions, personas, rather than a gigantic model that knows it all but it's not specialized enough to anybody, right? It requires both, right, to solve a big problem, so you can like research something going to the external LLM such as Perplexity or ChatGPT or anything, but you still need to apply that knowledge to what it means to me. So that latter part, what it means to me, is what we solve.
Jake Aaron Villarreal: And with a company that comes in and wants to get the value out of that model that's going to be trained specifically for them. Now, what's, what's the timeline look like if they wanted to take something like you have, your technology, train it up on them and then deploy it and get value out of it? Are we talking weeks? Are we talking months? What's, what's it look like today?
Suman Kanuganti: Yeah, it's typically 8 to 12 hours uh initial training period that um uh we start off with. And what normally happens is you know, by the time we are onboarding a customer, the customer already knows what personas they want.
Jake Aaron Villarreal: It's getting to that point now, right?
Suman Kanuganti: It's getting to that point. Uh earlier, earlier it used to be like longer kind of sales cycles if you will and it's like, "Oh you know what kind of personas do I need?" But now, now uh you know the word spreads and um and people are like, "I I know exactly what personas I want. Let's let's go." Right. So the way to... again, the way to think about it is you would have to apply the persona concept to people to make sense and make it easy, right? It's not complex. What do I say that like technical deployment that is happening where you'll have to figure out all the different connectors into all these relational databases and put everything into one place to God knows, you know, what you are expecting out of it, right? So we take a much more like simplistic approach. It's like, hey look, this is a platform. Let's start with what is your best bottleneck? Well, my bottleneck is uh closing the number of cases. You know, I need to increase my demand 1,500 cases um and be able to serve the actual demand which is 2,500 cases, right? Okay, how many personas do I need? What is the bottleneck? Well the bottleneck is being able to research the case and then coming up with the evidence that is required. Well, that's where you deploy the persona, right? And then you go in and you deploy and you you scale horizontally. Even the big companies, it's it's no different. We start off with like you know, few uh high value personas for a CFO, for a strategy officer, for a compliance officer, or you know uh for somebody in the compliance team who is overwhelmed and working 80 hours per week and the compliance errors are still pretty high, like what exactly is happening over there, and deploy a persona you know, that exactly what a human is doing or maybe a set of humans are doing into that persona, right? Um so yeah, so it's pretty fast.
I mean I haven't kind of articulated what the platform does, but Personal AI has a memory layer which is more about taking the unstructured data that is highly relevant contextually to a functional role or to an individual right. Think a corporate strategy or a finance analyst or a product manager or a marketing manager. Like you know you can think any role basically.
Jake Aaron Villarreal: Mm-hmm.
Suman Kanuganti: Um and there is a... and then basically you connect the data sources to it, right? Your existing data sources from your existing workspace or you know like uploading the data, uploading the data sources as well. The second thing is the model layer, which is basically our Small Language Model. This is where it's a kind of an ensemble model architecture that we put together, you know, over the past 3 to 4 years uh that does has a LLM kind of hook as well on a needed basis, right? So it's not dependent on LLMs as much as you can still kind of back off to an LLM if you do not have enough context within the, within your own model. And there is an API layer, so every persona gets their own API endpoint. So, Jake AI can have your own API endpoint on the internet, if you will.
Then you have a platform which is, our Personal AI platform is designed to be a more collaborative platform like Slack. Okay, you have humans and you also have your AI personas kind of working together and collaborating with each other, right? So, a persona I like to say you have an intake persona for picking up the phone calls and a customer is like, "Oh, I'm in a car accident." You know, I'm trying to get an appointment. Okay, you ask all the questions and that's an AI persona. And from there it could, the persona could put together the likelihood of case validity if you will, before it goes to a discovery persona. Uh but the discovery persona, before it goes in a human could potentially approve it. It's like, "No, no, no, this doesn't make sense, this makes sense." You can modify it and then you can hand it off to somebody else and that... So that's the application layer. So technically you can build like agents you know, on top of it for multiple different workflows. And I would want, but I thought I would I would explain it.
Jake Aaron Villarreal: Yeah. No, it's great, and there's a lot of engineers and a lot of other investors as well as um other founders on, in the audience. So I think that's really great just to kind of give that foundation. You know, as you look in the market today, what's, what's the competition look like in this space currently? You you talked about there's a category now like "Personal AI." So how do you differentiate from those?
Suman Kanuganti: Yeah, I mean the competition is actually great because it creates markets. You know what I mean? Um, where Personal AI differs is: one, in this core approach that we are training your model. We are not pre-training a model and then renting it to you as a service. Right? That's a fundamental core difference. This platform allows you to train your own models that are internal to you, for yourself, protecting your IP, protecting your data and still unlocking the goodness because it's highly contextual to you, right? So that's like number one thing. Um, and uh you know there are a lot of approaches that are evolving with RAG over LLMs, which is all great, which is a you know, similar kind of component structure that we have uh, but still like a lot of reliance on the, on the LLMs. Uh so I think this is the first thing: privacy, security, specificity um associated with your data and feeling safe and secure to unlock the value of AI for yourself individually or organizationally, right? So it's like number one thing. Um it's like that you know, brand of trust that we want people to embrace, and uh one of the core principles of the company.
Uh and the second thing is, from at least where we see and what we see with our enterprise customers, majority of the feedback who are investing in you know, AI, some form of AI internally, the companies are, they all sort of work, right? So it's reversed. In other words, they're trying to figure out how the technology fits in, not necessarily "I have a problem, I want to deploy or I want to apply the technology to it," right? Um we go in, we start with the highest problem space and we can quickly deploy a persona pretty rapidly um with high levels of accuracy, high levels of precision. So that's the second thing. From a pure like technological moat, um we have greater levels of accuracy than what we see in the industry and the ability to process like multimodal uh inputs and multimodal outputs. Not just like text input and text output, including PowerPoints and the CSVs. Like we are talking about like financial use cases. Um so the level of uh precision that we are able to offer uh is pretty intense, you know, as compared to the context windows or context sizes of you know a typical deployment I mean.
Uh and the final thing, it just comes down to the scalability. The way Personal AI persona architecture is kind of designed is you don't have to get overwhelmed investing multi-million dollars to God knows what the value is. You start with the top level personas that is most valuable and you scale horizontally. You don't need 20 million vendors for your AI value, right? You can scale a persona for a highly compliance uh need that is only internal specific, or you can come up with a persona that is external facing, that is highly brand oriented and marketing uh you know, and where people can you know, talk uh to to a bot on the website as well, right? So it's a, it's the scalability is pretty rapid and we can scale to millions of people. And some of the contracts that we are working on is is just that, you know, being able to scale a model for millions of people because everybody gets their own model, right? So yeah, I think those are some of the key fundamental um differentiating factors where we focus on and where like other markets are.
And I think AI is is truly the, calm kind of a thing from where I sit. So it's not one or two people going to win. It's more about a multitude of tools that will be available in the market. And some of these like vertical-specific SDRs, like for example, where people pre-train SDRs and you rent it out so that way they can bring in you know, some level of uh personalized data. It could be good for like, you know, small, medium-size businesses. But the same SDR persona would not work for a highly specific enterprise company that has their own brand, their own language, their own products, their own services. They need their own SDR. Well, that's where Personal AI comes into play because you can train that highly customized SDR persona for yourself that has your voice uh and your products and your services. So it's all about personalization at the, at the core via Personal AI. So that's our core difference.
Jake Aaron Villarreal: Yeah. And you got the name. I love the brand. I mean, how it's easy to remember. You know, when you talk about um the the the product itself and you talked about the investment, what, what's the business model like? What, what do companies, if you can share this, uh you know, just high level. Like what are the costs? What are we looking at if we wanted to you know, train up five agents to to help in our company in our law firm?
Suman Kanuganti: Yeah. I think what I would uh mention on a podcast setting is that we are 10x more cost-effective than some of the other alternatives that exist from a pure uh token generation standpoint.
Jake Aaron Villarreal: Okay.
Suman Kanuganti: Uh but we do require you know a few number of personas before we sign an enterprise contract. And our business model is basically price per persona, right? Um, so you either, you know, deploy or recruit human capital or you augment your existing workforce with AI personas. So that's kind of how we see our business model.
Jake Aaron Villarreal: Yeah, I like that because it's very focused on providing value to the organization. And if it's priced out the right way, it's going to be more cost-effective than hiring people. If you just look at it costs $100,000 for this person and if I'm going to have an AI agent or model that's doing that work for me at, you know, 5x, and the cost is, you know, lower than what my fee would be to hire somebody, then there's some win-win in there that you can really drive on the economic side. So, really cool. Um, what's the, what's the biggest challenge right now for the company as you head into 2025?
Suman Kanuganti: Uh, noise. Lots of noise.
Jake Aaron Villarreal: There's lots of companies. Yeah, I hear you.
Suman Kanuganti: I mean, I'm I'm completely okay with companies. I think uh this is the first time at least uh through the times of me, I don't think even .com kind of saw this through, or maybe a little bit. But startups as well as incumbents are playing in the same field, right? So, speaking of uh leveling the playing field, I don't think the field is level. There is no leveling the playing field if at all. If somebody is creating their own field, you got to just come and crush it. Um so, so I think uh the majority of the challenge is how do we communicate clearly the value to the customers. So that way you know there are definitely like a lot of good players, but there could be bad players as well whenever there is hype right, to seizing the opportunity. Um so my focus remains really deep-rooted to solving the problems and creating value to the customers. This kind of area we would focus on, and also one of the reasons why we try not to play too much in the like marketing games you know, even for what it's worth like the wars between like Marc Benioff and Satya around Agentforce and just the marketing push and being able to invest like these crazy conferences to put together the parties to push a brand is all great. But unfortunately, I mean I would want to see the level of value that is being created as well at the same time when you're creating this level of marketing hype. Uh and unfortunately they don't match. Uh if I if I if I if, if I do see that it matches I would I would say so but unfortunately it's not. So uh as a founder, I am underwhelmed with like the kind of marketing that somebody can push, but at the same time not have, you know, have enough emphasis on solving the core problems and creating a user experience that um is meaningful to the customer. So I think those are the things that we would have to navigate uh between you know just um making enough noise to get a faster you know, yeah contracts and valuation and whatnot versus you know here, here like no, no customer is saying that they are able to solve their problems with whatever they've done with their AIs, right? It sort of works, which is useless. That's basically...
Jake Aaron Villarreal: Yeah. And just so I guess going back to your comment about Microsoft and Salesforce, Microsoft being from the position of you know, an assistant, a co-pilot, a strategy where you know, it can help you get stuff done. Whereas Salesforce has taken the the really hard decision of saying or position of saying you know it's about agents that are going to actually do the work for you and not kind of help you a little bit. That's kind of geared towards actually replacing people to a certain degree. Maybe not everything they do, but certainly doing some of the work that you can kind of set it off, set it and forget it and come back with some good results. So, um there's a lot of talk out there and there's going to be a lot of benefits and results. So, we're curious as anyone to understand where where it plays. Specifically in our space too, we're in recruitment. So if you can go out and you can do a search and you can do a qualification, have a persona talking to candidates that are qualifying them, responding the way you would, you know, looking the way you look and then being able to drive the results and, you know, process if you're getting hired...
Suman Kanuganti: For Personal AI, we have a persona that interviews people.
Jake Aaron Villarreal: Yeah. So I mean, I know we already know what's here. We're already testing different technologies and it's really cool to see how accelerated the process has been from, you know, the conversations that were happening 24 months ago to where they're at today. I think it's just um yeah, we were on the forefront of that. The question we always like to ask companies when we're working with them, because we help them grow and scale, is who do you need to add and who do you need to let go? And every company has those hard decisions to make. You need to add people, you also, you know, what's weighing you down. If you could add an agent that can take part of that off your shoulders.
Suman Kanuganti: Yep.
Jake Aaron Villarreal: You know, that's a great way to start, I think. So, anyway, really exciting stuff. Uh, if anyone wants to find you or find Personal AI, where do they go?
Suman Kanuganti: I mean, Personal AI is pretty straightforward. You type in "Personal AI" in Google or anywhere else, you'll pop in. personal.ai is our website. Uh for myself, my Personal AI personal domain is s.ai and that's our platform and it's also a communication platform. Uh so yes, that's where you can find me and uh almost like you can add me as a friend and then we can talk to me, or if you're interested, talk to my AI as well.
Jake Aaron Villarreal: Yeah. We'll leave a little room at the end here and we'll, we'll add in some clips of what it's like to maybe talk to you or talk to a different AI. But I really want to thank you uh today for coming on, Suman, and sharing your story, and really exciting to see where you're at and love to connect with you back for the next 6, 12, 18 months and see how things progress. Um and to all of our listeners for listening. I really appreciate your time. It means a lot to us that you spent it with us today. I'm your host Jake Aaron Villarreal signing off for now. I can't wait to catch up with you all on the next episode. Until then, Suman, the world. Take care. If you like what we're doing, don't forget to subscribe, leave a review on Apple Podcast or wherever you listen, and follow us on YouTube where we go behind the scenes to learn what it takes to be a startup founder.
.jpg)
.webp)
.webp)
.webp)
.webp)
