Building the Future of AI Retail Merchandising and Planning with Jeff Fish Intelo Founder

When you're at this stage of the company, everyone you hire is a critical hire, right? Whether they're in product or engineering or sales or customer success, whoever whoever we hire is critical. And [music] within Intel, we eat our own dog food, right? So, every everyone is using AI tools all day long. And if you're not using [music] Gemini or Chat GPT every hour on the hour and you're in sales, probably not going to fit here, right? Yeah. If you're in marketing and you're and you're not using [music] a bunch of marketing tools that are that are AI, you're not going to fit here. Hi, my name is Demetri Bonichi [music] and I'm a content creator, agency owner, and AI enthusiast. You're listening to the AI Agents podcast brought to you by Jot Form and featuring our very own CEO and founder, Idkin Tank. This is

the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show. Hello and welcome back to another episode of the AI Agents podcast. In this episode, we have Jeff Fish, the co-CEO and co-founder of Intello.ai. How you doing today, Jeff? >> I'm good, Dimmitri. How you doing? >> I'm living the dream. Always have to always have to live the dream. Um, so just to kind of kick things off, tell me a little bit, excuse me, a little bit about your background. How did you get into AI? Um why what gets you uh I mean what got you excited about it enough to be running in Tel? >> Um I've been in tech for 20 plus years. I started out in e-commerce then I was in technology in talent acquisition and for the last 10 plus years I was

really focused on China and AI in China has been growing exponentially since I started there. Uh I started a company back in 2015 called uh Chatley which was a WeChat management platform and uh ran that for about 5 years until we were acquired by Salesforce and then I ran Salesforce China for about 5 years and during that time uh my co-founder and co-CEO had a product in engineering Rupesh had started Intel and had it in uh stealth mode for a long time and came to me and said do you want to leave your your job in a big corporate structure at Salesforce and and come run this with me. Uh and you know kind of you run the business. I I run the product. So he he runs product and engineering and I run the business. And uh he convinced me over about a year

to come over and and get on this adventure with him. We had worked together in a previous life uh before I started Chatley. And it was an opportunity that I I couldn't say no to uh for a few reasons. one, I really wanted to get into uh agents that were purpose-built for a very specific problem to solve. And to Rupes's credit, this problem was one that really needed solving. So that's why we we're focusing specifically on merchandising and planning and why we're um building agents just for retail. >> Yeah. No, that that makes a lot of sense. I think um it's really cool, by the way. I mean, just out of curiosity, what's it kind of like just in general having like a co-founder um and coc? I'm I'm an owner of a company that doesn't have any co-founders. What is that like in general?

And um would you say that it's I don't know. I'm c just curious kind of from your perspective what it's like. >> So, it's the second time I've done it. Um the first time the first time I did it, it was pretty similar. I I ran the business and my co-founder ran product and engineering. And as long as you know the two leaders understand there's you know lines of you know one person focuses on one thing the other person focuses on on the other and you can collaboratively make decisions. It works well. Um I I do have friends that have started companies where there have been challenges with with two people making decisions but I you know is you have to find the right person. You have to have uh similar beliefs in in how you want to structure the company, what your what your

values are. Ground the company in in the values that you both agree on. Uh I think that you know we we follow a V2 mom structure which is similar to um what I had done at Salesforce that that Mark Beni off and Parker Harris had started 25 plus years ago. And when Rep and I were were thinking about coming together on this, he had a set of values that I hadn't seen. And I wrote down the values that um that I wanted the company to have and they were almost line for line the same. So they were they were very much fully aligned in in the vision of the company and and where we wanted to ground the team. And I think if you have that then you can be really successful. If you've got differing values and differing um perspectives, I think it can

be very challenging. >> Yeah. Interesting. That's a good point. I I do I have heard something similar. I've heard like and and you know just to kind of speak on a positive note towards your co-founder like um I think it's great that obviously you're able to set those boundaries and whatnot. So really cool. Um and also really important I think to note that you know you guys have to have a different skill set probably for it to make sense in some respect, right? Like having duplicative skill sets I've heard is like not the best idea maybe for like starting together like you know because it's like they're getting a chunk of the the [laughter] thing. >> Yeah. Um he is a a retail technology specialist. Um came out of the consulting world. He he led retail for Sapiant. Um I came out of the e-commerce

world originally. There was a period of time where I worked for him. Um where I was actually in a product role and he was his CEO role. Uh so I think that we have similar capabilities. Um, but I think that, you know, this product is his baby and and I have a really good skill set in scaling. So, between the two of us, um, you know, focus on the product and making sure that it it works great and I can make sure that we can get it out to as many customers as possible and make it successful. >> That's very cool. Well, um, kudos to you guys for for doing that. I think I think it's awesome. So, um, more kind of into the nitty-gritty of maybe what you do and like kind of the origin story, a little bit more of yourself. Um, you

know, I uh I think there is something really cool. Obviously, you recently closed like a $2 million uh funding financial round from Illuminate Ventures. Um uh they're kind of known for finding gamechanging companies and um they don't rely just on pattern recognition. How did that partnership come about and what was it like us pitching a specialized multi- aent platform to investors? Yeah. So, you know, we weren't going to raise um we were going to raise in probably another six months. And when I announced on LinkedIn that I was leaving Salesforce to to come run this with Rupesh, Cindy Padnos and Jennifer Savage, who who run Illuminate, Cindy had reached out to me. I I had known her from um my Chatley days when I when I was running Chatley, uh asked if this is something that they should be involved in. and and they're incredible

venture capitalists and incredible operators and and really uh Cindy I' I'd consider a mentor and when she reached out I said you know let's talk and and we gave her the overview of the product we gave her where we were we talked about our pipeline a little bit and you know um she said well we want to be involved now and and we you can you can always raise another round later so um they let it um they're you know they help us with decision uh whenever we're making big decisions as the board and it just was the right fit at the right time uh you know when you bring in investors it's really important that you bring in similar to values of leadership in an organization that those investors are not just providing capital I mean capital's helpful but also understanding what your long-term

vision is wanting to be a part of that long-term vision especially early investors >> sure and and illuminators 100% those people so uh you know as we scale up, as we as we bring in additional executives, as we scale our sales team, as we scale our customer success team and product, uh we we we rely on them and we'll rely on additional investors like them as as we do the next round to to to really build something special and build something last. >> Yeah. Uh that makes a lot of sense. So ju just to kind of go one step back, I I I do kind of want to know um one kind of what level I talking about scaling your team. How many people are at your company right now? Um and then secondarily, I'll follow up about um what you do a little bit

more in depth. >> Yeah, we're 30 now. Um Oh, wow. >> Mostly product and engineering at at this point. Uh we are we are heavily focused on on scaling up the product first, but we are also adding customers at a pretty good clip. Um and we we're uh globally distributed. Uh so I'm I'm here in New York. Rupes is in Dallas. We've got team members in California, in Italy, in Hong Kong. Um and in uh in in Kochi, India where our engineering product team sits. >> Okay. Very cool. And just to like I said uh hone in a little bit more on what you're doing specifically, how was it? I mean obviously you're pitching what you did um in that uh funding process. kind of tell me a little bit more about what it's like building this uh AI agents platform and more specifically how

you guys are different in the retail merch sorry merchandise uh kind of realm of things with AI because we've covered a lot of products I think AI agents in general are now a common term but in this specific space what's like kind of the the thing that you really specialize in >> yeah so you know there are agents for everything now and having left Salesforce Agent Force and and other plat platforms like that uh can can be an all-encompassing platform where you can deploy, you know, 30, 40, 50 agents if needed um on a single platform. What there are not are highly specialized agents for merchants and what's amazing about this space is if you're in merchandising and planning and allocation, there's all types of tools that exist. There's legacy tools that have been around for 20 plus years. Um you could be le leveraging

your ERP or your point of sales solutions. You could be leveraging newer, more modern tools, but almost all merchanders and planners, no matter what space they're in, whether they're in high volume retail, big box retail, or super high-end luxury, the merchants and the planners spend a lot of time in spreadsheets. And there if if you were to look at what their tool of of choice is, it's not the systems that they have currently, it's the spreadsheets that they've spent years working on. And when we when we talk to chief merchandising officers across some of the biggest brands in the world, they'll tell you that, you know, it's a lot of art and and and science and math to be a merchant and a planner and to to, you know, pick what um trends are going to be going into next year or where the product

should sit and and you know, how much should be in the warehouse versus how much in the store. You know, we're going into Black Friday week uh in a few days. Black Friday's going to be here and merchandisers and planners. >> Yeah, it's crazy. [laughter] merchandisers and planners spend um months preparing for this and making sure that they've got the right inventory and the right mix of um of goods for for times like this. And they do it a lot in spreadsheets. So our biggest competitor is actually spreadsheets. Uh but when we looked at when we looked at you know what are the problems that they're trying to solve problem number one is there's an efficiency problem. So if you're spending 50 60% of your time in spreadsheet you can't be a strategic thinker. you can't think about, you know, how do I plan for

the next season's trend? How do I make sure that I have the right inventory in the next three stores that I want to open? But if you have an agent that's doing a lot of that spreadsheet work and you're and you're engaging with that agent, that allows for you to really think strategically. So, that's number one on on the efficiency side. Then, the next one, which feels like such a simple problem to solve and should have been solved over years, is just how do you get the right inventory to the right store at the right time? And all of these systems and tools that are out there are very rigid. So they're not they don't have the flexibility to um look at different constraints, make them wider constraints or narrow constraints. They're a lot of them are ML that have been built over the last

10 12 years and they're kind of blackbox AI. Whereas if you deploy verticalized agents for very specific problems and you can reason with them, it's like you're hiring additional staff. And there are no organizations that have unlimited opex. I don't care what space you're in. You could be the highest in luxury retailer or the highest volume. Um, no one's giving you unlimited opex. But what you do have is you have flexibility in software. So this is really why we're we're leveraging our agents as collaborative intelligence. And that collaborative intelligence is if you're a merchandising financial planning team or you're an inseason allocation team or you're a strategic planning team, you can add these agents and collaboratively work with them the same way if you were to go hire five or six PhDs that don't sleep or eat or take breaks to work alongside you to

to make your teams better. And that's kind of where where we're positioning our agents alongside these teams, not in replacement of and not um instead of the products that they've already got, but really as additional teammates that will deliver much better results and and much more efficiently. >> Interesting. Yeah. So that that that kind of touches on some things that I think I'll probably get to back to later in regards to the replacement or not. Um you know, actually maybe it fits more right now. What do you think is kind of the main um concern or kind of questions you get asked then as somebody who's in in this position? You know, you're you're building a a tool where someone would maybe say like, "Oh, are you trying to replace jobs?" Like, do you get these questions from potential clients? Do you get these uh

did you get that question from potential investors? Like how do you approach this whole situation with uh because it's it's a hot button issue right now. >> Yeah, it's not unique to us, right? It is a hot button issue. I think that, you know, this this notion that the a the agents and the AI are coming to take everyone's jobs are very overblown. Um in fact, if you look at the MIT study that came out over the summer of the Fortune 500 companies that have um that have implemented AI, 95% were failures. So, uh, that this change management of >> going through this new industrial revolution and that's really what it is, you know, since it's the three-year anniversary this week of Chad GPT. Um, it really has changed how you should look at your entire technology stack and there isn't an organization that isn't

doing that. And this idea that you're going to replace all of these people with agents is not realistic. and and and in and in our space when we think about the the art and the science, right? The the agents are really good at things like pattern recognition, are really good at repetitive tasks, are really good at workflows. These are areas that if you're a merchandiser and a planner and you went to school for this and you built your career on this, you don't want to be doing those things. You want to think strategically. You want to think creatively. You want to work with your product teams and your design teams and your finan and and your finance teams to to build out the plan that's going to make your company more successful. So the agents aren't replacing the agents are actually helping these people and

these these teammates do what they really want to do. And and in our case, we we've had great results. So uh you know we have higher sell through we've got uh more efficient teams. We've got teams that are um thinking, you know, two, three year plans down the road where they haven't had time and they were only doing one year planning. So the the agents are are here to help, not here to replace. And now are is that the case everywhere? I think you know in some cases where you know you've got some low-level customer service um a agents maybe that um or low-level customer service people that are just an answering questions like where is my product? I think agents are really good at that, right? And if that's what your job is, then maybe you want to level up and become a higher

higherend customer service individual that's doing customer success for for a brand and and helping upsell and those types of things. But in our space where there is this art and science, I don't see agents ever replacing. I see agents really enhancing um because those opex budgets are not expanding rapidly and I don't think they ever will. I think that's a fair point, honestly. Like, um, if then logic is big with AI right now. I think it's they're incredible when it comes to if then logic solutions. And what you just said was, I think, very important that like it's more art or it has such a high level of art in there. Like there's no reason to assume that there's going to be anything uh possible that is capable of overcoming the I guess the hum human human the human aspect. I almost said humanity aspect,

the human aspect of it, right? like um I would kind of just say it's interesting because maybe retail and or let's just say unitbased operation outfits could be considered in the realm of mainly deterministic right with if then but it's interesting um that you know you call that out because >> yeah I mean I wouldn't have maybe put it outside of the realm of possibility that it could you know just replace a lot of that uh those jobs because it I don't know unit based type things with like brickandmortar businesses sometimes to me just feel very like in out inout amount of you know objects but there's more to it right >> yeah there's more to it there's definitely more to it you know what what's the right product mix for a region in the southeast versus the northwest right what's what's the trend analysis

that you know is hot [snorts] on Milan fashion week that you need to think about for your next set of styles an agent can't doesn't think creatively like that. And uh you know, Elon said that we're going to have this utopian society eventually where no one's going to have to work. I don't believe it. Um I I I think that, you know, if that's the case and and it happens, great. I'm I'll be happy to sit on a beach while while an agent does my job, but I don't I don't I don't see that happening in my lifetime. Um and I don't see it happening in the space that we're focused on anytime soon. >> No, that's that's a fair point. I think the Milan call out is fair. or like Milan Fashion Week like you know I asked my girlfriend what the heck's going

on in that case and I have no idea. So I and honestly um even people who are in fashion I feel like it it's is this sort of like ethereal um who like it just takes a certain humanity like like we were saying that just doesn't um necessarily translate to anything agentic. But going going a little bit more into the strategy thing that I do find interesting. How have you helped um or how have you found though that reasoning models and the different improvements in AI have maybe helped on some level of strategy to an extent? Um and is that the case? Is that something you're finding? Now, it may not be the final final strategy, but at least helping get some of the leg work done for a human so that the human can spend more time on the more highlevel. Uh >> yeah,

for sure. So, so for us the reasoning is key to success. So, you know, when we first started showing this to to to our customers or at that point prospects, a lot of our reasoning was very technical, right? It was based on the data. We were looking at historical sales data. We were looking at um current inventory and giving really technical responses because we were tuning our reasoning engine. and we're integrated with all the LLMs, but really you have to have your own reasoning for for for this because you're rooted in in a in a customer's data. And what we were finding was that the the feedback was this is too technical. You know, these are not technical people. These are not engineers that you're working with. These are these are merchants and planners and and if it doesn't feel like you're talking to a

person and they're not giving you really good responses, the the adoption will be low. So, we've spent a lot of time working on the reasoning engines to have real business value conversation. So, it really feels like Demetri, like you and I are talking that when you're talking to the agent, you're able to to reason with it and have a conversation with it and then find out why it's doing things and and and then we started to see, you know, that that adoption scale really changed. So I think that you know the difference between kind of a legacy ML model or even um ear early stages a few years ago of of LLMs as reasoning gets better and as you can interact with it and and for us because we're um working with enterprise customers it's got to be rooted in enterprise data that's you know

safe secure and scalable um sure >> that's I think where the the the the magic happens and where individuals go Wow, this is great. It's not going to take my job, but it's going to make my job a heck of a lot better. And and I think that that reasoning element it it for us, it's the key to success. And if we didn't have it, I don't know that that our agents would be very successful. >> Yeah, it's funny. You said when the magic happens, and I immediately thought of like the the chat GPT like consistent output of where the magic happens. And I guess people do say the phrase. It's like a running joke on my team. It's like it's like when do people say this? I don't think they type it, but I think people do. >> People do say it. Yeah. [laughter]

So, >> um, yeah. No, that's that's uh that's interesting. I guess let's uh dive into what you just talked about uh a little bit more kind of by um I I've been thinking in in this respect. There are so many different avenues that one could start off with as a business owner um whether it be in retail or otherwise um to bring in AI agents into their into their company, right? How do you kind of approach specifically working with different clients on what be is best for them right when you're when you're um starting to work with them? >> Yeah, I I think I would take a step back and answer that from where I was up until earlier this year. Um when I was at Salesforce, Agent Force kind of looks at the key elements of what Salesforce offers, right? CRM, customer service, sales,

marketing, commerce. and you there you could have applied you know an agent the simplest one is you know a customer service agent right so if you're if you have 3,000 people doing customer service in a in a contact center somewhere and you need that that example that I was giving you know where's my package or you know I've got a broken item that's an area where you would tackle or if you got a sales team and you wanted to you know give give the sales team an update on on their on their records or their analys is that's a really good place to, you know, help your sales team do follow-ups. Like, for example, send an email 3 days after you interacted with with a um a record, right? That the person would forget to do, but the agent will just do it in a

workflow, right? In our space, our our um area and the purpose-built agents are only solving very specific problems. So, we've got, you know, now 20 or so agents that are basically 20 very specific problems that need to be solved in a in a limited vertical. So, I I I I tell our investors, our team, and and we're grounded in this and our all of our customers that we're about a foot wide and a mile deep, right? So, we are Yeah. So, that's that's kind of the focus for us. So, when we're talking to a luxury brand or a high volume retailer or a specialty brand, it's about, you know, do you have a merchandising and financial problem? If you do, let's dig into those problems and here's an agent that's going to solve that problem. Or, do you have a line planning or an allocation

problem? Let's dig into those problems because you probably have two or three people or 20 people depending on the size of your company working on those specific problems. An agent can come in and solve those problems. And with that with our agents because we're built on MCP with the Ato protocol, our agents talk to each other and then they're going to talk to other agents. So if you're a Salesforce customer, a Service Now customer, Workday customer, whatever that is, and those agents want to interact with ours, they can. And if you fast forward 3, four years from now, if you're a CIO, you're most likely going to have agents from multiple brands like Intello and many others. you're going to have to orchestrate those agents. They all have to be on the same protocol and then they have to work with your human teams. So,

>> sure, >> for us, it's just those very specific problems that we're looking to solve. We'll continue to look at problems that need need agents to solve for them and we'll and we'll build those agents and scale those up. But it's not, you know, tell us your problem and we'll build an agent for you. It's we're in this very narrow space. We see these problems based on what the space is telling us. And then if we see a pattern of customers that are telling us the same thing that we haven't thought about, then we just build it into the pipeline. >> Hm. That's interesting. I uh I like the phrase an inchw a inch um wide and a mile deep. I've heard the opposite be used for people who have interest miles wide and an inch deep. Like >> that's not that's not good for

what we're doing. No, but I I think it's very important uh from an agent standpoint because I never utilized that phraseiology, but I think it I think it is important to to note that like agents I don't think and AI in general maybe to this extent as well. Like I don't think they're actually in a position to do things exceptionally well generically. >> I agree. I agree 100%. >> Yeah. Would you say that this is one of the bigger like misconceptions that people have when you first start talking to them as clients or in general and you know what any type of conversation about AI? You know, I went to uh an AI retail event, the first first one of its kind called Retail Club a few months ago, and I was amazed by how many execs in retail couldn't really tell you what an

agent does that, you know, they they they know they need to be an AI. They probably have, you know, a chat GPT enterprise um account or maybe they're using Gemini through um through through through their Google paid uh structure. But if you really dug into, you know, are you guys deploying agents in your business? They don't really know what that means. [laughter] And I think that's most of the world. You know, we live in this space. You run a podcast that's talking about AI all the time. I I I run an AI uh agentic business, but 97% of the of the world is just trying to, you know, figure out their own business and be successful in it. And they know they need to they know they need to implement AI. I think there is a top-down expectation from CEOs in the Fortune 500 and

beyond that if they don't start to deploy AI properly, they're going to be in big trouble. And they're and you know there's the saying you're either an AI company or you're going to go out of business in the next 5 years. I agree with that. I think that every company is going to be an AI company. But if you dig down beyond we need to do this and you start to go, well, what is it do I need to do? There's a lot of questions. And that's why I think that MIT study was was so profound because there was a lot of dipping their toe in the water and not a whole lot of, you know, setting goals, setting KPIs, and then deploying what you need to deploy. Now, I think that's going to change. I think it's rapidly changing now. I think, you know,

you hear every day in the market, are we in an AI bubble? I think we're in an AI bubble. I think we're in like the second inning. Like the first inning was three years ago when Chad GBT came on. Now we're in the second inning. But now it's about, you know, deploying AI use cases and for us agentic use cases that solve real business problems and deliver real real ROI, not this is cool, so let's try it. And I think that's >> talk about the talk about the bubble a little bit more. I I want to hear what your thoughts are because there's a lot of different potential bubbles people keep throwing out there. What do you mean by it? Yeah, I mean if if you if you you know turn on CNBC at any second of any given day, you'll hear someone talking about,

you know, we're in an AI bubble because of Nvidia, we're in an AI bubble because you know, the build the build out around um data centers, we're in an AI bubble bubble because we don't have enough electricity or we're in an AI bubble just because valuations are too high. I think you know whenever there's a huge technology transformation, there is a curve, right? and and we're just at the beginning of the curve and there's a long way to go. But I think at the at you know when we get to the seventh or eighth inning to keep keep it in a sports analogy I I think we're going to have some amazing companies that are that are leveraging AI properly. >> Yeah. No, that's that's fair. Um do you think that there's an actual like evaluation bubble or like it's inaccurate? I I've heard some

theories that basically like there's like a double sort of counting maybe of value due to like you know flagship models are being funded and then the company's using the flagship models as their you know bread and butter so to speak like is is getting funding too. >> Look we're we're an aenic application player right and I think that the the open AIs of the world the anthropics of the world are very highly valued and are private. Yes, [laughter] Nvidas Nvidas of the world are the most highly valued company on earth. Um I think that you know they're they're highly valued because they're delivering amazing technology and I think the next phase is the application layer which is always comes after that right and there will be applications that do a great job and there will be some that don't and I think in any technology

cycle the ones that do do a great job and solve real problems and and have adoption will be very successful some will not make it and some will get acquired. And I think that's just, you know, that's the nature of of tech. But I think that we're not in a bubble. I think, you know, maybe there are some overvaluations. I mean, there, you know, if you're a publicly traded company and your your numbers are ridiculously high compared to your your revenue, then sometimes that needs to come down, but that's not I'm not I'm not a financial analyst, so I personal. >> You know what? That's a humble response. I'm not a financial an analyst, so I don't bother with that. That's fair. No, I like that. That's that's a good point. Um, we open up, you know what, a lot of people, there's a lot

of conjecture going around. I interview a lot of people, a lot of positive things, but no matter what we're saying, it's mainly conjecture on these things. We don't know what's going to happen, like practically speaking. >> Yeah. Um, >> I I will say, you know, it does seem like though, um, you kind of get that it's very specific, uh, what you're doing and that you're trying to make it as specific as possible. Well, going back to that last point, I think with the commentary, I don't think any of the major LLM players are going to get that deep at any point in time. Um, they're going to attempt to be generically good very well. Um, so I I do like this route you're taking, which is like a mile deep, um, inch wide, because it's just it's not the same thing. So, I do think

there is inherent value in focusing very well on something to this effect. And I am curious then does that mean like your ethos what is your ethos and like long-term vision of this company in this specific industry? >> Yeah. I I think that there's a lot of retailers in the world. There's a lot of organizations that need to move products and and plan plan inventory. And we think that we've got a really long runway to grow and I think that you know our we are still very much in the early stages. We're heavily focused on luxury and specialty. Um, you know, we're going to move into big box retail, hard goods retail, uh, grocery retail. We're going to focus, you know, beyond US and Europe and and get into Asia, Apac, where I've spent last 10 years of my life, um, focused on. And I

think, you know, there's just the the TAM and the SAM are both really big if we just focus on this and, you know, revisit that in three years. Uh when we've when we've grown significantly over the next three years, we'll worry about it then. But for now, that's where we're focused. >> The old TAM, Sam. Very true. Um yeah. No, I I like that a lot. I I I think your approach makes makes sense to me. >> [snorts] >> Um, another question I kind of have is, um, you know, this last year this podcast, uh, launched. It was called the Yeah, agents podcast. Practically speaking, no one the heck, no one in the heck knew what an agent was. >> Um, it was all chat bots if you recall about a year ago. Kind of moved pretty quickly. Um, what did you kind of do

as a product to I guess make a a shift uh or not shift but like a a coming onto the market as an agent product and trying to educate people on that immediately because I don't think it was like fully explained at the time like six months ago like feels short but a lot that that was a long time ago. I don't know if it was quite maybe fully in the market or how did how did it feel kind of jumping in because I'm sure there was prior to the six months ago there was probably a level of pre-planning, you know, just like launch a company. So, how did you kind of prepare to get into that realm when it was kind of a new market? >> Yeah. So, I mean, we we just came out of out of uh stealth in January, right? Repeat

was building for a long time. Um, and had some had some early adopter customers, but was really in stealth. Uh, and I I think the the sales forces of the world with agent force and and service now and and Microsoft with co-pilot kind of laid the groundwork for enterprise organizations at least to understand what agents were. I think if you were to you know rewind two two years ago. >> Yeah. And I think you know leveraging that that marketing capability of of of the ma of the major tech players helped. I think if we had come out of stealth a year before and we said we have purpose-built agents, nobody would have known what it was. >> But >> now, at least where we are and and where where we're playing, everyone's heard of agents, but like I said earlier, just because they've heard of

them doesn't mean they really fully understand what it means. Uh but that that level of of you know knowledge of agents are something I need to talk about and think about that is that is widely um adopted within within the Fortune 500 Fortune 1000. >> Yeah, I think that's a good point. Um I'm not saying I did a bunch of the leg work either, but like you know podcasts like mine, other people kind of like just like talking about it all the time. I'm sure to some extent like uh there there's like just there must have been the general thing you were able to kind of jump on with what other people were saying to to kind of make it make make it clear. But like in your first attempts to sell sell it right like um one-on-one and with calls was it was there

any difficulty there? Was it like easy enough to kind of explain? I'm curious uh kind of how that was with customers at the beginning. >> Easy to explain. a lot of fear amongst merchants that these are here to take my job. And that the you have to you have to really dig deep and show that the agents are not here to replace you. They're here to help you. Um I think that's still happening today. I think I think we still have those conversations a lot. You know, I'm sure I'll have one later today. I'll have one tomorrow. I have one Wednesday or two Wednesday before we break for [snorts] Thanksgiving next. But I think next >> Yeah. So I think that's that's still here. Um, but you know, you have to demonstrate that your product, whatever product leaders you're talking to, whatever product organizations you're

working with, um, are are, you know, helpful as compared to there to replace or, um, you know, skynet you. >> Would you say that you feel like this is something that is going to go away from a fear-based standpoint? And when do you think that would be? >> I think that technology transformations throughout history will have fear amongst certain people. You know, >> um >> that's good. >> When the cars came, they were if you were a a buggy driver, I'm sure you were very fearful that the car cars were going to take your job, right? Um so I I think historically that's just the way it is. Uh, I think that that'll happen here, too. But, you know, it'll become so widespread and so standard that that will dissipate over time. I can't predict on when that will change, though. >> That's fair. No,

I think that's a good point. I mean, I remember uh watching the movie uh I Robot. Just kidding. Um, no. Do you remember that movie with Will Smith? >> Yeah. Yeah, I know the movie. >> A lot of people kind of just seem to have this like funny like movie level concern about various things, whether it be AI or aliens or whatever it is. And um sometimes it's like I wonder what would happen if we just remove that like backlog of Skynet, the iRoot, like all that kind of stuff from people's mind and like how they would approach it practically, right? Um I think their heads would be a little more clean on the subject, right? Like um it's it's almost like a Frankenstein analogy. Like everyone always is uh funny and like when you're a kid, you're afraid of uh Frankenstein and then you

realize it's actually called Frankenstein's monster. And then you get to an age where maybe you read the book and like the movie just came out so that's why you read the book and you go like oh the monster's the good guy >> right [laughter and clears throat] you're like oh Frankenstein's the monster that's the whole bit got it um so yeah no I think I think a lot of what we are struggling with right now is is coming from sensationalism on the news and or previous like movie experience that just kind of makes us concerned like practically I'm not like I think the dystopian type movies are fun, but that's all they are. They're movies and we're not until it happens. Like I don't think it's going to be something that actually occurs. But um just talking a little bit more as as a founder

um you know when you uh are are just going about your day-to-day life, is it uh difficult to try to plan out what you're going to do from a hiring perspective? Um, if say for example, I'm not sure like what level you're still entrenched in hiring. Um, but like when you're hiring people, I know you're an early guy, I figured. Yeah. Like when you're hiring is like individual capabilities with general AI uses usage like top of mind when interviewing? >> Yeah, 100%. So, um, this is my second rodeo. I I have friends that have done this several times. Starting the companies in tech, you have to be a little bit crazy. You have to be willing to work, you know, 80, 90, 100 hour weeks. Uh all of those things are very real and they've been documented over and over again. When you're at this

stage of the company, everyone you hire is a critical hire, right? Whether they're in product or engineering or sales or customer success, whoever, whoever we hire is critical. And within Intello, we eat our own dog food, right? So every And if you're not using Gemini or Chat GPT every hour on the hour and you're in sales, probably not going to fit here, right? If you're in marketing Yeah. Yeah. If you're in marketing and you're and you're not using a bunch of marketing tools that are that are AI, um you're not going to fit here, right? If I if you're an engineer and and you've never used cursor and you've never used Anthropic or Claude, you're probably not going to make it through the first interview. That's just the way we are, right? Uh and that's really important. So, you know, we have Slack channels that

are just around what tools we're using and we're measuring the tool usage on a weekly basis. I literally just sent out, it's Monday morning, I sent out some tool usage information this morning to the team. So, yeah. So you literally are tracking whether individuals at your company are properly on a daily basis utilizing AI enough. >> Absolutely. >> That's pretty >> absolutely across multiple tools. >> That's pretty good. >> Across multiple tools. Yeah. Um if if that's not in your DNA or you're not willing to embrace that in your DNA very quickly, this is not the place for you. >> Yeah. You got to get with it or get out of the way. Honestly, um if you're if you're kind of especially at an AI company, right? I think nonAI companies, you know, I I have a little bit more understanding of way of of

why they may or may not be, you know, a little bit more cautious in respect to like, oh, I don't want to like overly do it, but like no, I I think for what you're doing, I I think it makes a lot of sense. So, I don't think it makes much sense to be anything but cutthroat about or maybe not cutthroat is the right word, just like aware of where people on your staff are lacking, right? You you're young, you're lean, you have to kind of build out a team that is around the ethos of your company and also honestly the the projection of the future. Like there's there's been a phrase that has gone around since AI came out which as a platitude but seems like you're trying to at least go along with it where I do think a lot of companies will

say this but not actually follow through on it. They'll be like >> well AI won't take jobs but if you're not willing to utize AI you're going to you know have opportunities taken away from you for people who who do >> and you're practic you're practically actually uh measuring it which I don't think necessarily a lot of companies are doing. Yeah, we're practically measuring it. It we it's in our interview process. If someone if someone says to me in an interview that they're not interested in in leveraging these tools, they don't make it to the next round. >> That's a wild statement to say. I would not I would I would pretend to be fair. I would I would at least pretend be like ah like saying I don't want to like interviewing at an AI company just sound a little um >> yeah we're

l we're fortunate enough I haven't interviewed anyone that said that. the lack of self-awareness would be crazy in that context. Um ag goodness that's crazy. >> Um well on a personal note kind of what what is some of the stuff that you do on a daily basis tool-wise? I know you talked about cursor you talked about a couple couple other tools. What is your favorite like tool to use for personal use to get more work done? >> Well, I I was using Gemini Pro 3 this morning for about two hours and the newest the newest model if you guys haven't used it is awesome. So that was just this morning. Uh I use for our presentations, every presentation feels like a custom presentation to any one of our customers. I use Gamma and I was building. >> Yeah. So I I was literally building a

deck this morning for a meeting right after this um in Gamma. Uh I use that all the time. Uh I you know we use cursor as I mentioned. We use we use claude. Uh, I I can't say that any one is my favorite because I use use them so often, but I think the one that I probably use the most is is Gemini just because we're we're a Google shop. >> Okay. Yeah, cuz you're Okay, that's fair. Um, from a actually from an agent standpoint, not to give away the secret sauce too much, but like internally when you're creating agent tools, like do you like bounce between models a lot? Are you like constantly testing with like the engineering team like what models work for what things? >> So because we built our own reasoning engine and we have our own workflows, we we're actually

integrated with all the major foundational models and >> Yeah. So we're not we're not any we are not focused on any one model. Uh we've got customers that want to use DeepSeek on Azure because it's cost effective, right? We're okay with that. If you just want to use Open AI because you've got an enterprise open AI contract and and you want your tenants just focus on that, we can do that too. But generally our our multi-tenant strategy is Gemini, Claude, O Open AI. Those are the three that we're plugged into that are standardized, but you know, we're we're we're agnostic on on the foundational models. I think you were mentioning earlier about, you know, which one's better, who's going to win. I think it's I don't think it's win or take all. I think that, you know, there's room for for all all the models

to be really great and um you know, we're lucky that we're in a time where there are so many great models that are out there and I think they're just going to get better and better. >> What do you think is like the best one for uh research right now? >> I mean, what Gemini just released, I think, is is is the best that I've seen. Uh I I play with all of them personally. Uh that one's really damn good. I really have been kind of pushing Gemini for not to say I started the move. No, I'm just kidding. Um, no, but then when 2.5 came out, I was like shocked with like how solid it was, but how like little traction I feel like in the market it it got from like a a love standpoint. You know, there was not much of a

I guess uh you know explanation from from people as to why. And I think it was just because like the Bard and early Geminis were pretty bad. >> Yeah, they were bad, >> right? They were bad. >> They were bad. Yeah. Um, and Catchy BT comes out with five, right, after basically saying this is going to change the world, you know, and it goes really bad. >> Yeah, it's not good. It's not It's It goes It's not good relative to the expectations. Um, >> relative to what Sam had had tweeted about. [laughter] I think that that's what the you know, you got to set the expectations properly. >> Yeah, he kind of went a little overboard. And then the funny thing was that afterwards GPT 5.1 is actually solid and was a >> probably what five should have been. And [snorts] the funny thing is

that I remember seeing a tweet from somebody saying like, "Wow, GPT 5.1 is not getting a lot of love. They really didn't do any press for it. Um, it should get more love." And then Sam Olman literally replied to the guy and said, "Better this way than the other way around." LOL. I was just like, at least he's self-aware of that um of that failure. Um, well, just to kind of close out the conversation on on a, you know, positive note, obviously we're talking about LLMs and that's that's great or whatever, but outside of general LLM, right? Like there's different tools that are out there that are really good to help you that are augmented by AI. I'm a big fan of a tool called um, Text Cortex as well as a there's one called Crisp AI, which is a um, noise cancellation AI tool.

So basically, you can be on a call and it will transcribe your call and will reduce the background noise significantly. It looks like there's like a 0.5 one second delay, but if you're like on the go or whatever as in meetings on job sites and stuff, it's like so worth it. Um, plus, you know, I think meeting transcribers are like the most underrated tool on the planet right now for people who are in industries that are just like, I don't want to record my calls cuz security. I'm like, "Doesn't Google have those secretly for like the first 30 days?" Anyways, um but anyways, uh like people don't know this. Like, I'm pretty sure Google like has to have it on their books for some like weird reason. I'm almost positive. Maybe don't don't quote me on that. I'd have to I have to look this

up, but I'm pretty positive like it's in their cloud storage. Um so, read terms and conditions. Those are my two favorite tools that are not directly like a model. What are your what's your favorite tools right now that you use that aren't like a model >> that are not out that not our own product? Um I love gamma. You know I mentioned it earlier. I you know [clears throat] as I think I'm a little bit older than you. I know I'm a little bit older than you. I I remember having to to to to build powerpoints and you know get pixel perfect presentations 15 years ago. And the idea that you know you just create a prompt and you have this killer killer deck in like three minutes. um is awesome and I I you know I don't want to keep pushing that particular product

but that's one that that I love that's got nothing to do with I don't even know what LLM they're plugged into. I don't think it matters. >> No, I I totally agree. That tool is really good. I'm I'm a big fan of it. I've used it for um presentations presentations I've shown you uh [laughter] and it uh it's it's really incredible how you're able to just take an idea and kind of run with it so quickly. So, um, yeah. Is there anything else that you'd kind of like to plug? Uh, obviously your own product, duh. But outside of your own product to kind of take, um, everyone into the into a good ending mindset wise, as we close out this episode, I I would say that, you know, for anyone that's listening to this or watching this, you got to look at various agents that

are specific to what your problem is. Don't don't, you know, try to settle on one big platform. You know, I keep going back to Salesforce because I used to work there and they're partners and friends, lots of friends there. You know, they opened up the Agent Force uh partner network, which we're going to be a part of, and there's other players that are going to have them, too. Uh don't try to solve all your problems with one thing. I don't think it's possible. And that's what I would recommend for anyone that's listening to this or watching this. >> H totally fair. I like that. I think um it's a good comment. That's that's fair. That's a good way to look at it. So, last but not least, where do you want everyone to go in order to check out what you're doing? >> Intell.ai or

reach out to us on LinkedIn or how there's lots of ways to reach us, but intell.ai is the fastest and easiest way to reach us. >> Absolutely. So, with that being said, everyone make sure to go to intell.ai and also make sure to leave a like on this video on YouTube, give us a review on Apple Podcast, Spotify, that includes you, Jeff. We want to get it out to the most people possible and share it on all of the different platforms that you can imagine because we're just trying to make people know what the heck AI agents are just like the cool one that Jeff and his team are building over there at Intella.ai. That's intell.ai. Thanks so much for watching and we'll see you in the next one. Peace. Thanks, Demetri.