Episode Transcript
[00:00:00] Speaker A: Foreign.
[00:00:11] Speaker B: Welcome to the Victory Show.
[00:00:14] Speaker A: Hey Victors, welcome to this episode of the Victory Show. If this is the first time you're joining us, I'm Rachel Lee with Bestseller By Design. Our founder, Travis Cody is the best selling author of 16 books and we've had the privilege of helping hundreds of business consultants, founders and entrepreneurs write and publish their own bestselling books as well. Through that journey, we've discovered a fascinating pattern. Most businesses really struggle to break past the seven figure revenue mark. On this show, I sit down with some of the world's most successful CEOs, leaders and business owners to uncover the strategies they used to scale way past that mark so you can do the same. So get ready for some deep insights and actionable takeaways that you can implement in your life and business. Starting now. Today's guest is Joe Tibelani, the founder and managing director of FreshRiver AI, a leading generative AI consultancy specializing in delivering mission critical business outcomes through agentic AI frameworks, workflow automation and human centered design. With over 25 years of experience at the intersection of technology design and systems thinking, IT has helped Fortune 500 enterprises and government agencies transform their organizations through advanced AI solutions that generate emotional, ethical and economic brand value. A recognized thought leader and New Jersey Innovate 1002025 honoree, Jyoti is a sought after speaker and advisor on Gen AI Strategy, System Level Transformation and ethical AI Design. Her work bridges data science, engineering, social science and design to shape sustainable, inclusive AI ecosystems. She also serves as a Columbia University School of Professional Studies AI advisor and founded the AI Transformers Leadership Council, an executive committee of over 100 enterprise leaders accelerating adoption of next gen AI. She is deeply committed to mentorship and equity in AI and through Fresh Rivers fam, the Fresh River Apprenticeship and Mentorship Program and her advisory work, she empowers women, early career professionals and executives to upskill and lead confidently in the generative AI era. Jyoti, welcome to the show.
[00:02:13] Speaker B: Thank you for having me. I'm really excited, Rachel, for our conversation.
[00:02:15] Speaker A: You've had a long and incredibly impressive journey in tech and design. What led you to founding FreshRiver AI and what, what was the moment you knew it needed to exist?
[00:02:24] Speaker B: I think chronologically I'd say 2018, probably even 2016. I had this feeling that I need to do more. I'd spent about 20 years or so in the corporate world. I actually grew up riding the telecom wave, wireless and wireline. Two critical technologies that have come about. One was Ethernet which was increasing the bandwidth for communications and then Fiber optics, you know, being a software engineer in my early career and then doing product development work and then marketing towards the tail and all in senior roles. You know, the telecom industry was basically consolidating. Usually every industry that has its first sort of takeoff goes into an S curve and then kind of tapers off. So I was finding less innovation taking place. And that's really what excites me. I love building things when there's a spark of something new that comes on the horizon and of course, connecting the dots between different aspects of design. So I'm a product thinker. Bringing in now they're calling a Z shape. It used to be the T shape and now it's Z shape, where you have multitude of skills, you're deep in one or two, but then you also have the emotional intelligence piece, which is getting even more critical. A I knew I needed to leave corporate and I fell into happily, natural language processing, a branch of AI and conversational AI. And the crossroads I had was either to go to blockchain or natural language processing. You know, thankfully, the universe pulled me into the, the AI space. And so I started to do independent consulting work for these Fortune 500 companies. They would bring me in as a consultant, say, hey, we've been testing and playing with these chatbots and, you know, one of 2000 vendors in the space that had these platforms. They said, we're just not getting any value out of it. Can you help us? So that's really when I started to see a couple of things. One, AI is very different from any other technology we've encountered, whether it's mobile or SaaS or cloud. For a key insight that I gained in that AI is not a deterministic technology, much as software has been, you know, input, processing output. You press a key, you're going to get the same thing over and over again. AI was not that. So that's when Rachel I realized that there's a disruption and innovation and what we know, and it's completely different from anything we've any of us have encountered. The other key insight I got was a lot of IT teams were tackling AI just like they always have. Like, you know, technology is a hammer looking for a nail. And this is coming from a technologist. So, you know, that's kind of, you know, my, that's how I grew up. And I love my engineers because I, I am still one. But they needed to be a shift in thinking. There's a paradigm shift in multiple areas and that's where everything changed for me. And I knew that we needed to bring more than just capitalism or money, which is important. I'm a capitalist for sure. But if you think about economic, ethical and emotional intelligence, what I set out to do was I'm going to build emotionally, ethically and economically intelligent AI. And in order for that to happen, emotionally, ethically and economically intelligent humans have to build AI, if that makes sense. So let me pause here, Rachel, because I could answer your question in a very long way.
[00:05:40] Speaker A: Yeah, absolutely. I completely agree. It's sort of like, you know, the inputs do affect the outputs and even if the outputs change, I really like that you noted that it's really important that the people who are training it are thinking about it holistically. And given your approach on doing it ethically and sort of thinking about it from an emotional and economic value perspective, having that three part lens is really probably what has led to a lot of the success of Fresh River. Now Fresh river has been around since 2019 and that was long before Gen. I really hit the mainstream. Tell us about those early days of building the business and maybe how that evolved after ChatGPT really hit the scene in a meaningful way.
[00:06:16] Speaker B: Yeah, I love this question because everybody's like, you were talking about this before everything became like cool and hot, right?
[00:06:22] Speaker A: Yeah.
[00:06:23] Speaker B: So, you know, in my early days, it was a slog. I was a team of one.
I had done marketing the last couple of years of my corporate life. So I had a fantastic partner that I brought in where I said, all right, let's start doing email lead Gen. A couple of reasons. A, I was trying to position this consulting capability which I call strategic AI consulting. It's a strategic assessment framework. But I learned a lot. So over the next three years, I spoke to over 400 enterprise decision makers. I would take eight to ten calls a month because it was new. People were kind of sort of hearing chatbot and conversation like it wasn't mainstream yet. So I was getting a lot of interest and people would get on. These executives would get on and say, you know, we are taking your call because we want to learn, we just don't know. So spending a lot of time just educating. And so from that education I learned it was almost like primary research. Like, you know, how much people, do they actually know about AI? Do they really think about, you know, I was testing the waters, but not closing as many. I was actually getting all my projects through other consulting companies. They were subbing to me and because they said, hey, our client needs fill in the blank, they didn't have job descriptions this is the real fun part. They still don't have many job descriptions because a lot of people are using these models not to generate job descriptions. But people don't really know what kind of skill sets they need, right? So that was really amusing for me because the recruiters would say, hey, I got three lines I got from the hiring manager. Are you interested? I can't find anybody else. They actually used to cry on my shoulder, these recruiters, because they said, we can't find anybody else. So I said, okay, A, the demand's there. B, when I would take the interviews with the hiring managers, they would say, we didn't know what to write in the job description, but you're who we need. Can you come help us? Because, you know, I was learning with every project. And then I finally realized that there's a framework of thinking. So those were the first three years. And then I decided a couple of things. I said, okay, number one, there is a demand. Number two, all these opportunities are coming through other consulting firms. I'm going to go direct and through a lot of places I've been speaking at since actually 2019. I was part of this very large, this organization where there were enterprise leaders in there. And through that, a Fortune 10 head of innovation pulled me aside and said, hey, because I was running the special AI kind of council for that organization. And he said, we need help. We've been trying to do this pilot with this, a vendor. It's not working. Can you help? But we need a bigger team. So I actually did, just in time, hiring. I knew people in my network and I said, okay, I'm going to sort of lead the direction, the strategy. I had engineers, I had testers. I was training women, by the way, since 2020 on this technology in the middle of the pandemic, right? What a great way to spend time than to learn AI. And so I figured out how to teach other people what I do and of course, pull in engineers. And that was the start of it.
From that point on, started getting more and more work. Now getting to your point around 2022, and I, and I do, I'm a big believer on the universe aligning things. You kind of put it out there and kind of happens, which is the only way I explained to people how I got here.
I was consulting or doing professional services for this technology firm who I met, actually, that first direct client that we got. And they loved the way we approached deploying services, doing the use case extractions, things like that. And while we were doing, I was doing that Work with my team and training my team on that platform. They were the number one vendor at that time. I met a another gentleman who's now my co founder. He left corporate, he had built the largest financial trading infrastructure, he had worked on AI and automation for the last 15 years. And we both saw the same thing happen.
So if you think about these, so number one, the models before ChatGPT where all the vendors who were building their platforms or conversational AI, the models weren't good enough, they weren't delivering. Like if you think about, I'm not going to say her name but she might wake up. Alexa, right? Or Siri, you know, they're all listening, they'll wake up. They never got into becoming enterprise grade and we knew why and so we said now the technology has gotten really good. Know these transformer based algorithms, they're open source models that are becoming as good as these large models. They weren't back then, they are now. And we said, you know, for our enterprise clients these larger models are not going to work because of regulatory security, privacy reasons. They wouldn't put their golden family jewels of data into them. So we knew that was an issue. So we set out to build our own agentic AI platform and we looked at algorithms around the globe that would do what we needed it to do to become enterprise grade. So we effectively created our own platform and training open source models that are raw but putting all the guardrails around them and training them on the use cases that we're extracting. So in other words, if you think about these larger foundational models, they're like a big giant hammer that you're supposed to use and try to force on every little nail or different kind of nail that you have. We built precision AI, a surgical tool that goes in, we define the problem, we go in and we're so confident, Rachel, of what this, what we've built, we contract for the business outcomes. So we work with C levels and we go, okay, we're going to extract your use case, we're going to train the engines to create your AI agent, deploy it where you need and we're going to commit to your business outcomes. So we do value based pricing. So that's the three things that happened after. So if I divide up the six almost seven year journey, first three years was learning, actually doing hands on work and discovering as the tech was changing we could actually build something ourselves that meets our clients requirement and we deliver in 90 days which is unheard of. So we're getting a lot of the projects that are failing in the generative AI space. And what people really love about us is that we're educating them first. So these are all the senior executives who don't know how to speak about this subject yet. Their boards are like forcing them to say, you got to the gen AI, there's a FOMO effect. And we've got now a fantastic team of senior executives formerly from the industries that we are serving and they're helping shape these use cases and their engines. And of course they're taking us into their networks. So we've kind of got this sort of ecosystem built out that is building its own momentum. So I sometimes say we're like the antithesis to foundational models. What gives you owned and gives you rented intelligence that you can't really own to building owned intelligence. And these models then are owned by our clients. So we make our money through the professional services, the license of the platform because they can't train the model without it. And then per transaction that that AI agent performs, we get a percentage.
[00:13:26] Speaker A: Yeah, you mentioned earlier that you brought on a co founder. I'd love to hear a little bit about how you found them and the process of integrating a co founder.
[00:13:34] Speaker B: Great question.
He's a very senior executive from the banking industry from a technological standpoint, but he loves hands on coding so he never lost that capability. So you know, usually when you're rising through the ranks, you kind of let go because you're going through, you know, management.
But he said he used to be so bored. He's one of those prodigies that just think about. Yeah, I mean he has a PhD, he has education background, but he just loves creating new things. So. And I love defining the problems and we meet in product. So what happened was it was a great fit because I didn't want to go back to hands on. I did not keep up with that. I went more towards, I want to find the right problems to solve, define them in a way into a product. And you know, we're, we're both from industry, we don't really need coaching or guidance on what to do. We know what needs to get done. And so we've just been moving at very high speed, so very complimentary skill sets and we're just building on top of each other's skillsets. And of course the third part of this whole story are the senior executives from Domain. These are domain leaders from healthcare, life sciences, mortgage banking, financial services that are coming in and validating and helping shape. So it's really, it's a beautiful Thing we're not building out. You know, my aim is not to build out a traditional organization. I believe we're entering an era where people who have built industries over the last 20, 25 years, who know where the bones are buried, where all the problems are, I can quickly go find it. That's really what's making this very successful. And we're moving together.
We call it a flywheel effect.
[00:15:20] Speaker A: How did you two link up and how did you sort of figure out he was going to come on as the co founder as opposed to a hire? Tell us a little bit about that piece.
[00:15:30] Speaker B: So that the technology vendor that my team and I were doing pro services for, that tech vendor, happened to bring now my co founder in as a consultant for the issues they were having with their platform and for very large clients. And I still remember this day, it was every 2022 we were both part of this organization called iMasins, which is a data center consortium. And you know, I come from telecom. So for me that was like I was coming back home. And so I joined the New York chapter and we were at the meeting that was being hosted there and I got introduced to this gentleman when we started to talk and we weren't actually on the projects together, but we would meet up every time I'd be at the office and we started to talk through this. He was helping us with one of the projects that we were working on for the vendor for pro services. And then ChatGPT hit and he was telling me the challenges he was running into using that platform.
And we both kind of realized that when we saw the gen models and ChatGPT pushed out into the world, we kind of knew that was coming. The transformer models were out there. But not all vendors jumped on it right away because they had already invested their money in the pre transformer based technology. So I just said to him, I said, you know, we'll never, we, we, we haven't been able to solve mission critical problems with the models that we've had so far.
I think we can train our own. And he said, yeah, I think so too because we both saw the problem coming at the same time and we actually had a client. That was the beautiful part. So we had a client and we got to work with a proof of concept.
Then we did the next one and then we did the third one, a huge pilot and we realized we were onto something because the feedback we were getting from people. So it was a happenstance. It was a very joyful, happy meeting. Like, you know, when you have you know, atoms that collide because it's so random. And serendipity. It was serendipitous. I did not. I didn't know how I was going to go build it. I just knew that was the. That was the destination. And he said, yep, I think we, I know how we can get there. So just got drama? No, nothing. Just got right into it. Like, Edmond was so natural.
[00:17:48] Speaker A: What does victory look like to you today?
[00:17:50] Speaker B: For me, it's a daily thing because now that we built out the platform, we've trained models for those specific industries. Every day we get on a call with a prospect or a lead or a client, warm because our senior team is bringing in introductions. A look on their face, Rachel, when we show them what we've built and what it does, their eyes sparkle and their jaws drop. That's victory for me. That was always. Our goal is to build something with AI that has not been done before and we haven't been able to do before. So I'll give you an example. One of our leaders on the mortgage practice said to me, you know, every loan in this country costs between 10 to $13,000 just to close a mortgage loan. Just to close. So, you know, people like you and I were filling out this form, they're paying people to do that. And I said, what if we drop that by a third? He said, you know what that would do? Think about the affordability of housing. You're bringing the cost down. So the big thing we've got in our minds, Rachel, for if you could talk about victory, is taking cost out of the system, making things more affordable. So when we look at use cases, we're not automating the current processes. What we're really doing is doing a complete analysis surgically to figure out which use case, which area we put this AI agent of the brain in. If we solve this problem, there'd be a cascading value creation downstream. So we found those entry points and say, that's the problem we got because, you know, there's so many problems to solve. Any given enterprise has an average of 200 use cases. 200. And you have to find the right ones that will unlock value not just in place, but downstream. So to me, a watching leaders say, oh my God, we've not seen this before. We're not getting this from our current vendors because everybody's being bombarded by AI consultants, AI startups, big tech coming at them. Everybody's coming at them, but they're not seeing outcome. I'm sure you read the MIT study that just came out that says 75 to 80% of all the pilots have failed, and we know why. So when we arrive there, that's the one part. And then we tell them it's not just the part that we're going to fix for you. We're going to make sure that it makes you more money, because, you know, everybody loves making more money. So Victory has many definitions, Rachel, but it's about really solving problems that humans have not been able to solve because we haven't had that tooling and the way of thinking in order to do that.
[00:20:27] Speaker A: If you could give your younger self one piece of advice, what would it be?
[00:20:31] Speaker B: Don't be afraid. Don't be afraid to speak up. Don't be afraid to go ask. I did a little bit of that. I don't think I did enough.
[00:20:38] Speaker A: I get that. I think that's an excellent piece of advice. So often we doubt ourselves or don't speak up. So I love that. And, Jodi, this has been an absolute joy. Thank you so much for joining us on the Victory show.
[00:20:52] Speaker B: Thank you so much, Rachel. Great conversation.
