Brian Bell (00:00.736) Everyone, welcome back to the Ignite podcast. Today we are delighted to have Peter Satali on the mic. He's the co-founder and CEO of Sorcerer, which is building what he calls a the autonomous execution layer for global physical goods and trade. Thanks for coming on, Peter.
Peter (00:14.825) Yes, thank you, Brian, and really appreciate the time and looking forward to the conversation.
Brian Bell (00:19.522) Yeah, it's good to see you again. So I'd love to start off with your origin story. What's your background?
Peter (00:24.459) Sure. Yeah. So I grew up in San Diego and always was very interested in computers. And so I started coding when I was twelve. And my first project that I did was when I was sixteen, which was a mobile food delivery app. And so I was in class and said there should be some way to go ahead and order food and at the time, right? It's twenty this was back in twenty thirteen. when door dash was just getting founded and everything like that. And so I created this app and got a couple different restaurants. And but the thing is at the time I I 16, who's really gonna fund someone that young and do they drop out of high school? And so I continued doing that up up until college. And it just what really sparked though my whole interest in in entrepreneurship. And that's really what started during during college. I I did a company called Religio and it's really what was the foundation of me going in and starting different businesses. And so that's what was was the original aspect there and and has sparked kind of my whole worldview of using tech to solve problems.
Brian Bell (01:38.35) Yeah. Which was required back in twenty nineteen. What did that teach you going through that journey from founding a company to actually going through an acquisition?
Peter (01:47.185) Yeah. So when I created the company, I was thinking about first off opportunities of where tech is very antiquated right now. And at the time, churches was a very interesting opportunity because we really didn't have in some cases they didn't even have websites. They at the time they also really didn't have as much on the payment processing. And so the the first business of the whole software site was called Church Deposit. It was a payment solution processor for churches. And we started going to these different churches. And what we found though was the issue for a lot of these churches was less on the payment processing side. It was actually with the declining membership numbers. And so if we looked historically back in like the 1950s and earlier, the church was there where people go to make friends, network, do business with each other within the church. And today there really isn't that additional type of community. And so we pivoted business to Religio, which was doing essentially CRM systems. We had also a mobile app. We had still continuation of the donation platform, but being able to improve the overall community within these different churches. And so we were able to scale it up and then sell the business. at the time when I sold the business, it was back in twenty nineteen. So this was I think timing wise, you know, maybe I I should have waited a little bit longer in terms of twenty twenty one when there was a huge valuation increase, but sold the business and I what I really wanted to do, I guess, was because I've been doing it for quite some time, like d when I first started in college and then a little bit after, and I I wanted to try something else and and shift over to a different industry. And so I thought, okay, you know, I think it's it's time to to sell the business. And so immediately after the sale, I mean, for for me it was just it was a nice early win and and I wanted to be able to use that as a stepping stone to go off to do another business down the line and and have it as an aspect of a win that I could potentially show to either new hires that hey look I've been from the start to the end and I could do this for for my next business. And so I really learned a lot and it was a great journey.
Brian Bell (03:54.146) Yeah.
Brian Bell (03:57.519) So what did you s and then you went through speedrun eventually with the with this the sour sorcerer company. Was was that was it did it predate speedrun or and then you applied or did you kind of come up with the idea during speedrun?
Peter (04:03.568) Yes.
Peter (04:11.738) Yeah, so just in terms of the a career path, so right after I sold Religio, I joined a venture capital firm called Human Capital, which I was a principal there. And Human Capital, the the two leaders, Bars and Armand, what they looked at, if you look in terms of the last ten years, the whole thesis was most of the unicorn companies about and their and this is just their kind of rough numbers, like 90% of the unicorns come from technical founders. And so I I was talking with Bars and I was like, I don't know. I think he really bar Bars was the main partner at at at Human Capital, who's the founder of the fund. And we were talking and I said, look, I I really want to build another business. I'm not sure if I want to kind of do the VC route. And what I think was very strong human capital, was being able to be in such a strong network of these top engineers that you go ahead and be able to be connected to and potentially start business with and really learn the process of of also to recruiting. Because one of the unique value adds that human capital did was being able to go in and essentially similar to how there's a talent agency for athletes and movie stars, create almost a similar type of dynamics but for top tier software engineers. And so being able to go to startups, and why would someone want to take money from from the fund? Well, it'd be because you have this whole internal recruiting team that would go ahead and actually support the the the the the startups and be able to recruit the engineers. And so I learned a lot during that process of potential interesting business ideas, but also to the recruiting side and be able to work underneath these very, very top tier investors who they invested in SpaceX, Snowflake, Brex, Andrew, so very strong in terms of their their track record and be able to understand that's at the table. And then immediately after that, so I was there and then left to do basically an import business. So I was importing different medical products. And so that was actually the early foundation of thinking through hey, this could be something opportunity here because I was importing a products and
Peter (06:30.474) I was focused on supply chain. And I there's so many aspects of antiquated systems. I said there has to be a way to to make this be more smoother. And so I created Sorcerer. And at Sorcerer we're building the fully autonomous supply chain and and we started the business a little bit before applying to speedrun and I I applied to the speed run and we were accepted in officially the the program started in July of of last year and and we just hit the ground running on that end and my my co-founder and I we've been able to to scale it from and that we we did our the demo day and a seed round and we've gone from the beginning of the year we're about three hundred thousand a monthly orders. now we're doing about one point five million a month in orders. So
Brian Bell (07:24.589) wow. Huge, yeah.
Peter (07:25.757) Huge increase and now we've officially passed over two million ARR as well. If we look in terms of our take rate, I know we'll get more in depth about our business, but speedrun and A sixteen Z has been really tremendous in terms of what they've been able to to give us on guidance and there's so many aspects that they've been able to help us on. So we're really appreciative.
Brian Bell (07:33.688) Yeah.
Brian Bell (07:47.257) So much I wanna unpack there because there's just so much content that we just breezed over. one of the things that was interesting is you were on my side of the table for a while. How what are your thoughts on being a VC versus being a founder? Kind of what are your because you've been both now, yeah.
Peter (08:03.06) Yeah. So a couple things. I think a VC's could kind of and the way that I I liked how Barge kind of explains it, you have two two options on what you can do, whether if you want to be a founder or VC. The the the situation with the VC is good is that you can go ahead and you can make a lot of bets, and then you if you one of them becomes this huge outcome, then you can be able to go in and get that aspect and it's also diversify a little bit. Of you don't to put all your eggs in one basket on that thing, versus if you're a startup founder, you're really all in and and and that type of thing. So I think those are kind of the the core on that side. I think there's different types of venture funds. At the time, human capital was early stage, and that's vastly different experience than if I would be like a late stage, like a series C, series D. Because those late stage ones are just a lot of it is financial modeling, it's looking at other aspects versus the early stage, you're really, really trying to dig deep. And understanding the person. What's their background? What what are the type of things that they've done that's really incredible for for whatever they are in in terms of their their career path? And being able to understand that I think is is is a lot different than going in and and doing doing late stage. I would say the main aspects that I I liked about VC was that you'd be able to constantly go up to date of all the different stuff that people are are working on and be Energize, like wow, this person could really go and change this one industry and those type of things. I think though, for me as a founder, I think it gets to a point where I would like to be the one that's and again, obviously it's it's it's a hard journey, but being going in and actually impacting the future. And I think VC though is to its credit, right, that it's needed to before these these entrepreneurs to be able to go in and and start their business. But at least for me, I really like being more on the operation side and stuff. at least for my my friends that have Gone that I've known that have done like because me was a little different because I went from startup founder to then VC, then startup founder again. For those that have just started with VC and then never did startup stuff, sometimes it's a little bit of a learning curve because it's it's you know, you're you're it's not so much of a calmer time. You know, there's a lot of ups and downs, a lot more hours. And so I think those are some of the main differences on that end. I think the startup life is.
Peter (10:25.595) More fun, but it's also, you know, you're it's a journey. And so it's a different aspect for for for people depending on what they're interested in.
Brian Bell (10:32.788) Yeah. And we'll we'll get into speedrun a little bit, maybe a little later. But I want to talk about Sorcerer specifically because, you know, the the growth is really impressive. You know, going from you said three hundred K a month to one one and a half million a month. That's like a five X increase. And how do you measure that? What do you guys do? What problem are you solving? How does it work? Yeah.
Peter (10:50.715) Yeah. Sure. Yep. So I'll just step back a little bit and just explain a little more about the supply chain. And so within the supply chain, there's a lot of these fundamental issues. And there's difference in pricing, there's quality issues. And I think a good example of this is something as simple as a pen, right? So someone's importing a pen for one dollar a case, and these are let's say, for example, these are large scale distributors. So these are importers that are doing Hundreds of containers from overseas into the states. For for those type of people, they they might be buying at $1 a case and there may be exact same similar factory than another company that has a more sophisticated sourcing system, be able to get it at 50 cents on the dollar. And so if we look at overall supply chain, not only is it very manual type process, but it's very opaque aspect. And so what we're trying to do at at Sorcerer. Is that we're trying to create a new supply chain. And what I mean by that is supply chain 1.0 has all these different aspects of a large amount of opaqueness. There's a whole bunch of other other aspects in terms of delays and quality issues. And at source, what we're trying to do is use AI agents to improve the entire ecosystem. And so there's different steps of the supply chain. The first step, which is what we focus on, is the sourcing side. And sourcing within procurement is going in and finding factories. And so with the example with the pens. we have AI agents that will go in and find every single pen factor in the world and reach out to them, negotiate various different platforms as well as also use AI agents to actually call to get a hold of these factories and in a ton of stuff on the research side. And so if you're able to do that in real time and something as simple again as the pen example, and you're able to reach out and find every single factor enrolled and do this continuous process and compare the freight and the tariffs, you have a huge systematic advantage.
Peter (12:48.793) Be able to actually lower your costs. And so what we're what we're doing again. So step one is the sourcing. The next step is the diligence side. And so let's say you find two or three different factories. How do you know, right? If the pricing's slightly different, maybe one, two percent different in the pricing, which one do you go with? And it's very tough because fundamentally, I mean, if these factors, so yeah, we can for sure take your your volume, but if they get a bigger order comes in, they're gonna prioritize that order. So there might be delays, there's also could be issues with quality. And so we have agents that will do full and background checks on the factory, be able to actually autonomously book third-party auditors to visit the factory and be able to do that end-to-end. And then the third aspect would be what I would call the the just the freight side. So comparing the freight and be able to better understand if there's any delays. And the fourth step is the purchasing. And so if we look at overall on the supply chain, how how it tentatively works, is that the more volume you do, the cheaper the price. And so for a particular customer, to like one of our product categories, as I say, plastic cups, for example, we'll be able to go in in real time find across the whole globe what is the cheapest factory in the world that it that still has good, strong quality, good delivery time. But we're able to do is through export records, we're able to find other buyers of the same type of product and aggregate the demand and help everyone save. And so that's how we've been helping our customers, which are Fortune 500s in terms of end customers, about 10 to 12% savings. And some of our customers do billions of dollars of transactions. And so we have started on a smaller scale with them, but being able to really prove to to them and get more and more volume through through our platform because of it. And so to answer kind of the the order aspect, the way that we define it is the the amount of volume going through through sourcer, right? So the beginning of the year January is 300,000 And then in terms of for last la I guess for this end of this month we'll be at one point five mil monthly orders. last month was one point two mil. we're able to get a take rate on that. And and so that's how we we make our money. And now we've passed over two mil N AR, which has been really quick growth from the beginning of the year.
Brian Bell (15:01.934) Right.
Brian Bell (15:11.894) It's incredible. I mean, I guess a a question a lot of ECs think about when they invest is like, why now? Why has anybody built something like this before? Because it strikes me as something and think I I I probably asked you this question when we first met, like, why didn't somebody build this 10 or 15 years ago? Or maybe they tried and they just didn't succeed. And what is it about now that just made this really attractive for you to go build?
Peter (15:21.394) Mm-hmm. Sure. Yep.
Peter (15:33.354) Yep. So I think there's a couple things. I think if we took at the history of I I guess in terms of supply chain tech, the way that I look at it is the the way that I I look at it is a couple things. there's marketplaces. So marketplaces people have tried and unfortunately they don't work. And the reason why a lot of these marketplaces don't work, and and to the fullest extent Is because there's risks of circumvention and a lot of these these factories don't want to be on this type of platform. And the reason why is that if you're a manufacturer and you're going on this marketplace, the issue is that you're you're gonna now be competing with more and more people. It doesn't benefit you. So a lot of them don't wanna be be on those type of platforms, and obviously and this is what I'm referring to as business to business. A lot of stuff you with like the Alibabas of the world, they're business to consumer. So they're going in and that type of thing. But these are for distributors, and just so the audience understands the distributors what it what it means in the context, these are large-scale importers where they warehouse the goods. And so, like a food service distributor example would be one that imports and let's say plastic cups, and then they sell to a bunch of different restaurants, for example, and with all the different napkins and everything. So that's what a distributor does. And so
Brian Bell (16:31.437) Right.
Peter (16:55.047) The hard thing is, the if you're constantly having to compete on price and you then create this marketplace and it becomes very, very inefficient, your your pricing goes down because the my the the supply sire supplied makes money in the picnic. And so they're incentivized to not share the pricing and stuff. And so that's once a yeah.
Brian Bell (17:14.412) Yeah. Perfect competition is like perfectly, you know, perfectly destroys your mar margins, basically.
Peter (17:19.931) Yeah. So that's that side of the the equation. And then the separate side is there are tools right now, and I know we'll get into a little bit later about some of the stuff with AI agents, but there are aspects of procurement tools which are good, but it fundamentally doesn't solve the tenets. And what I mean by that is a lot of these tools they'll go in and help save time. They do manual processes for like PO processing and other stuff like that. And so what what's interesting right now is that when we reach out to the factories, the factories have no idea that it's an AI agent. And so what we're able to do is in real time essentially create an internalized marketplace where we understand the pricing power better than anybody else and be able to go in and and and also on top of that. So that there's there's two type of questions. We're essentially going out and doing that the the full process. So we go in and we can be able to negotiate in real time and just an up to date pricing for everything better than anybody. But then we we we're able to accumulate so much data of other importers across the globe that we can then go to these individuals and be able to then offer them very, very strong pricing that they'll go ahead and and bring their volume over to us. And so it creates a network effect because the more volume we do, the better, the better we're able to go in and and do a bigger negotiation leverage with the the the factory. And so that's how we've been able to do it, and that's why one of the reasons why we've been able to win so much business. Is because a lot of these customers are tired of being s sold up a false promise by a lot of these tech companies where they say, yeah, we're gonna help you save money and they they don't actually save money. And what we're able to do tangibly is if someone's spending like some of our customers spend billions of dollars a year and we're starting on a smaller scale in terms of their their volume, but to go in and say, give us a chance and let's see what we can do and be able to tangibly go in them and offer them lower prices at 10 to 15 percent, which is a huge amount of savings. And direct RY. And I think one other aspect I'll I would mention about this is a lot of these workflow automation tools are looking at the wrong aspect. And so if someone's spending, let's just say $10 million a year on their procurement team, a lot of these these workflow tools are like, I'll help you automate it and you can maybe cut half of the team. And so yeah, you save five million dollars on a billion dollar budget though of of of annual spend.
Peter (19:43.204) What we do is we look at different and say, look, you're you're currently spending a billion dollars on on all your materials spent annually. Let's actually go and attack that. That's actually what you care about, more than this smaller amount of your procurement budget. And that's what we're able to do for our customers.
Brian Bell (19:54.382) Yeah. Pretty incredible. So it's it's almost as if you've built, you know, how this was done before Sorcerer is you might call some sort of broker, right? And say, Hey, I need to buy a million pens or whatever it is, a million cases of pens, dollar a case. and they'd go around and try to figure that out for you. And they say, Okay, I got like three suppliers and here are the prices, which one do you want to go with? And blah, blah, blah. But now you can sort of build a marketplace using autonomous AI agents. Which effectively creates a marketplace and creates a marketplace platform without needing to onboard people on both sides.
Peter (20:34.105) Yes, a hundred percent. And one of the things that that's very important about what we're able to do because of the v the the overall the volume aggregation, there's there's not a risk of circumvention. Because the big issue with brokers, right? If that's even a broker that's brokering transactions, sure, maybe they'll get paid on that one transaction. But once the two parties understand who's the factory and who's the buyer, the broker goes away. Yeah.
Brian Bell (20:58.03) Yeah, I'll just I'll just go around you in order from the factory. Yeah. Yeah.
Peter (21:01.302) And so what we're able to do that prevents those types of situations is because we do so much volume. And I'll give you an example, right? So if one of our one of our this is one of our our factories, when we started working, we had zero percent volume with them. Now over 50% of all the volume that that they get is because of us. And so part of the things that we're able to do is that factory, like so let's say there's one factory that they get they have a capacity of 50 containers a month.
Brian Bell (21:19.757) Wow.
Peter (21:30.911) Or we're able to do is very quickly be able to find all these different buyers and I grid demand. And the reason why and let's say one of those people that are doing five containers a month, let's just say, goes to the factory and says, Hey, look, I want to go direct, they're gonna say, No, the reason why you're getting a better price is because of all that volume. Because you're on you're being going in with with with forty five other containers that are being packaged with yours, that brings the cost down. And so it prevents those type of circumventions, even if they they they want to do something along those lines, it prevents that. But regardless of you know, the the the the customers don't even want to switch because the reason is we're able to help guarantee no delays. We're also able to make sure that the quality is hold to a even greater standard than the industry norms. And so we have a not in terms of up to date, we have no quality issues for any of the products and because we go above and beyond for the quality. And so it's those three aspects is really what has driven the growth and be able to help us win business.
Brian Bell (22:32.738) Yeah, again, it's like this story of like AI enabling this new sort of business model, which is an old business model. It's basically broker dealering, import export, or kind of marketplace. But you're able to leverage the labor of AI agents to build the supply and the demand and ensure the quality all at the same time. That's that's pretty incredible. and did you know it was gonna look like that when you started? Or like how is it how has the vision changed since you got going?
Peter (23:00.747) So I think a l a lot of the VCs are more accustomed to more of the SAS type model. And I I think that as we've seen in terms of the public markets, the SAS has completely been decimated a lot of the the stocks, because of what's happening now where people can go in and spin up some sort of SAS thing with Claude and you know, codex and and stuff like that. And so what Our viewpoint has been is that we think again that AI agents will continue to get more and more sophisticated. And so I didn't want to do the the whole aspect of going in and and and helping these companies where you're going in and and doing these workflow automation type things. I think that though though in the beginning, when we first before we got invested by A16Z speedrun, A lot of the the venture funds really wanted us to go more that direction of going more the workflow because it it's what they've accustomed to. And I think right now if we look at the overall supply chain tech
Brian Bell (24:05.294) It's how it's how you kind of underwrite things. So like it's a hammer looking for a nail. It's like I that's how I underwrite is your your seats or your volume. And I that's just what I understand. And without that metric, I don't know how to evaluate you.
Peter (24:15.249) Yes. Yep. And that's the thing. And so there's a lot of you know, pattern matching and stuff like that. And so what we've done and what has been our main driver of the growth is that we really understood what these customers cared about. And sure, the workflow automation is a nice to have, but fundamentally, what do they care about? They care about getting their stuff on time, good quality, and cost low. And those three aspects have delivered the the main the main value for for our customers. And so I I would say that right now we've been been able to to keep with our thesis. Our thesis has actually stayed consistent, even though we've had a lot of push and pulls when we first got started, because this was just such a unique way to approach it, that it was just fundamentally very different than what VC funds would have seen in the past.
Brian Bell (25:09.422) Yeah. Yeah. And that's coming from a former VC. and you know, as a fellow VC or current VC, when I met you guys, I was like, okay, yeah, this is a unique w thing at the right time and a unique way to solve it. And that's kind of what I look for when I make an investment is you're you're solving this age old problem that people have, right? Supply and demand, import export, quality control, low prices. Nobody's ever gonna want it slower at a higher higher price with lower quality, right? And How can I enable like how does AI enable that in a new way? and it was like, yeah, that was like a a no brainer. What do you think? what what's a popular idea right now in AI, maybe with agents or AI generally that you think is wrong?
Peter (25:56.155) I think I would probably double down on the whole workflow the workflow automation aspect. I think just expand a little more in depth. I think there's a lot of companies right now, so what was hot obviously for the longest period of time was B2B SaaS. Obviously we've seen what has happened. Now what we're seeing is a lot of the workflow automation tools that they're that they're gonna go to kind of industries and be able to help companies do workflow automation and I think the main negatives of that is that you're competing with a lot of the the labs itself. Like if we look at more recently, Claude has the new update where you can record your your your workflow and it will automatically then start doing it for you. And so I think in the early goings, like it like in in in terms of like last year or two years ago, like helping these companies, it made sense. But now what's gonna again, I I I
Brian Bell (26:44.002) Wow.
Peter (26:53.349) It's hard for me to predict what the labs will do, but I mean we have to presume that they see this whole opportunity here of all these companies trying to do workflow automation that, you know, why don't they do it themselves? And like OpenAI had the yeah. Yeah.
Brian Bell (27:03.576) Yeah. It's just kind of core to what they do, right? And yeah, and I have I have a couple of those in my portfolio where I'm like, oof, yeah, that's probably not gonna work out now. But they're probably not gonna get into what you do, right? It's too it's too complex, it's too thorny. I can't imagine anthropic releasing the the import export supplier.
Peter (27:23.715) Yeah, I mean it's again it's it's it's we do have defensibility there because you know we're we're the ones that handling these transactions, those type of things. And that's not to say too, I mean, I still think there will be winners in terms of the workflow automation and stuff. And I still think there's gonna be stuff where they don't really wanna do the hassle of switching over and and and those type of things. But I think broad base, what I think will probably happen is that there is gonna be a margin compression for a lot of these workflow automation. type type companies. And we've already seen it already with I I I'm sure when you go to SF you see all those ads for the AI SDR. And so they were charging you know a monthly fee for being able to do their you know the the AI agents and then what happens they became so many competitors because there's not as much differentiation between a lot of those AI sales aren't rep that the the margins did compress. So Then there's gonna be a little bit of that. And again, it's hard to predict what what happens. And I think there's also still a lot right now with the whole distillation aspect. I don't know if you've you've covered that yet, where you know, it's it's it's getting tougher for laps to be able to have a huge advantage for a short period of time, because I just to kind of explain to the audience, the way the distillation works is that you can essentially ask certain questions and you can see how Claude thinks. and how it responded that you can essentially create a replica, not not to the fullest extent, but be able to kind of create your own version. And so we've seen a lot of like more recently there's been certain new new AI competitors that come into space that have done those type of things. And so I think that's something that's also not as talked about, but it's gonna be interesting to to see what what ends up happening if you know that also impacts The main business itself in terms of the the the the labs itself, the frontier labs.
Brian Bell (29:22.786) Yeah. Yeah. They might be innovating their their way out of their their own margin, right? Right now they're printing cash, but you have Kimmy and and the deep seek models right behind them. And I think recently the Kimmy models overtook them in some of the frontier benchmarks for a fraction of the cost. So for 90% less, you can get the same intelligence. you can even self-host it and save a little bit more money if you have a, you know, a use case where you got to throw a lot of tokens through it. You can probably save another 30, 40, 50 percent hosting it yourself.
Peter (29:37.719) Yep. Mm-hmm.
Peter (29:52.131) Mm-hmm.
Brian Bell (29:53.027) versus using their APIs. So yeah, it's gonna be, it's gonna be, I think 2027 is gonna be a really tough year for the labs, I think. Because at some point it's gonna be like there's it's super intelligence, which is great. It'll it's gonna do a lot of amazing things for humanity and let's not discount that. but there's gonna be all these use cases where the Kimi three point, four point, whatever models are good enough and they cost a a buck a million tokens. and you know you know it's funny you mentioned the distillation
Peter (30:00.928) Yeah.
Peter (30:18.955) Mm-hmm.
Brian Bell (30:22.914) Back when Deep Seek came out, if you would have asked it who created you, it would have said open AI. You know, back when we had the Deep Seek moment was w last year, year and a half ago. yeah, it's amazing. So back to Sorcerer. So the obvious VC critique here is hey, isn't this just like a services or brokerage business of like AI? and I think, you know, I'll let you answer that, but I think that's a little bit of this like generational overhang I think we're seeing in the startup and V C community a little bit is we're
Peter (30:28.223) Yeah. Yeah.
Brian Bell (30:52.696) We're trying to apply the old SAS lens to to new new ways of doing business. And AI is sort of eating all these services and broker all this like human labor now.
Peter (30:57.793) Mm-hmm.
Peter (31:01.888) Sure. Yeah, I mean I think a couple things. I mean, I think i if we look at our our business, what we're trying to do is create a new supply chain. And so fundamentally, this is different than just a standard service business because we're fundamentally impacting global trade. And I would say that in terms of a lot of the the VCs are trying, you know, and it this is just in general, they look at it and they say, you know How are you going to go ahead and get recurring orders and stuff? I mean, these businesses and that we sell to these are fortune 500 in terms of end customers, and they're going in and consistently getting containers of plastic cups, pay you know, a lot of different napkins and those type of things. And so they're gonna continue to plate more and more orders on a consistent basis. And that's just the the nature of it. And so I think if we look at just SaaS as a whole, I think SaaS you know, has a lot. greater risk of being wooldered you know winded down more so than what we're doing. Because there's a lot of stuff fundamentally that what we're able to do is go in and consistently offer the better pricing and and guarantee the quality and delivery time, which no one else is doing. And so these things are actually, regardless of the economic conditions, people are going to still need aspects of supplies to can come into the country and it's it's has a lot more defensibility, I think. and and and it also does a lot stronger in terms of tangible value for society. I mean fundamentally if we're able to do this right, we're gonna be able to help the whole the the whole consumer base actually have lower costs of goods because we're able to go in and run this whole process and make the supply chain more efficient. And so that's what I would say and answer along along those lines.
Brian Bell (32:57.186) Yeah. Yeah. AI, it's just an another example of how AI will create this productivity boom. We haven't really felt it yet, I think. But you know, I think it took a decade to really feel c the economic impact of computers. And I think it'll probably take another, you know, five or ten more years to really feel how AI is really I mean, you're you're you're I mean, look at look at you guys, like like the value you're bringing, but right now it's only a like a small a small tiny fraction of of the potential future impact. speaking of, like what are you excited about in the in in the coming year or two?
Peter (33:34.715) For our business or for just general AI? Yep. I think right now we're on a very strong trajectory. We've been growing twenty-six percent month over month. And for us, we just want to continue that growth. And I think what's very exciting is that one of the Fortune Founders that end customers does over 60 billion a year in spend, and another one does over 10 billion a year in spend.
Brian Bell (33:35.884) Yeah, for you. Yeah. Yeah. Both. Yeah.
Peter (34:01.852) And we have a huge opportunity where on a smaller scale, we've been able to help them save about 10 to 12 percent of their cost. And so if we're able to go ahead and really deliver, there's a huge opportunity for us to expand quite quickly in terms of the volume. So that's that I would say on the more revenue side and what's what's forthcoming. But on the technology side, we already have the core piece in terms of the the autonomous sourcing. but what I think is Interesting is that on the on the more diligent side, we're expanding even because a lot of our customers they're looking for options outside China. And so we're able to actually pull a lot of information where we can find factories in Vietnam or Cambodia to be able to offer our our customers a a lot stronger in terms of the delivered price. And so I I think right now we have a very strong focus in terms of Southeast Asia. And so we probably expand a little more into Latin America and and kind of Mexico in terms of areas of getting more of the supplier intelligence than what we are right now. We're mostly very Asia focused in terms of the the supplier base along those lines. But that's what I would foresee. And and I think we've already seen in terms of our take rate it's gone from roughly beginning of the year from January to May, about 10% take rate. Now it's gone up to about 12%. And what I think is very exciting about what we're able to do is that as we get a stronger, stronger network effect, we're already seeing a stronger amount in terms of the take rate because we're able to go in and get more and more volume. And how we usually work with a lot of our customers is we we start with the stuff that is the more commoditized. lower margin type of products. And so it helps us get our foot in the door. And then once we've win their their trust, we can expand into more of the higher take rate type of products. And so what I think is exciting is that we can potentially be a situation where having accelerating not only on the monthly orders, but also on the take rate. And so that's what I would say is exciting as for us in in terms of the business and be able to execute the next eighteen months.
Brian Bell (36:28.162) What paint me a picture of the long term. So if you think ten or ten or fifteen years out and you accomplish the the f the vision and mission for the company, what does that look like?
Peter (36:37.312) Yeah. I think the things that I would say that would be a little bit different. I mean, there's a couple of things that when we get way later down the line, like eight, nine years, there's gonna be a lot more sophisticated aspect in terms of AI agents actually doing the purchasing itself. And so it's no longer going in and having a procurement person doing this. You're gonna have AI agents that are gonna be constantly negotiating both on the factory side and the buyer side. and so I think longer term, what what I see, a couple of things. I would see on us being able to in certain product categories having as essentially like an autonomous group purchasing organization. And so being able to go in and and help a lot of these different customers, in this case like food service distributors, all be able to get lower costs because of the demand aggregation and because of the constantly checking of the pricing of those type of things. I think for us, if we're able to get this right, we'll be able to go in and actually expand into other product categories in terms of more so along metals and other areas to be able to to completely dominate a particular industry. Because it's if we're able to go in and we've already done this a smaller scale, but to go in and continuously have the best pricing, continue to get more and more demand. Because fundamentally if we're able to do this, we're gonna become the main way to do a trade. Because you go in and how who has the best price? We have the best the best price, then people buy from us. we can then continue to get more and more customers and it just continues to become the main way of of doing things. And so for us longer term, we can be able to get billions and billions of dollars of transaction. And I think the the way that I would say like how we feel confident about this, I mean one of our customers does by themselves a 70 billion a year in spend. And so I I don't see a reason why we wouldn't be able to get billions and billions of dollars of transactions and and be able to really make this a massive company.
Brian Bell (38:32.524) Yeah, and I I recall this when we first met too, like I I think asked you like what the TAM is. It's like it's like what's the global trade of you know, like of goods, you know.
Peter (38:42.282) Yeah, it's like it's like thirty six tr trillions and I I when I was a V when I was a VC I never liked all the stuff where it's it's a huge, massive edges. Right now we're focused on, you know, a pretty category, but it is yes, you're right, it's trillions of a lot of a lot of time. Yeah.
Brian Bell (38:45.464) Yeah, tens of trillions of dollars of TAM, right? Yeah.
Brian Bell (38:52.546) Yeah. Yeah, you gotta like you got your wedge and you're kind of work your way up, but I mean it's literally tens of trillions of dollars, right? I mean that's that's a huge tan. And then you kind of start like the the way you're starting. let's wrap up some rapid fire questions. what did you believe with total conviction five years ago that you would now argue against?
Peter (39:08.053) Okay.
Brian Bell (39:35.766) Or inverse, what did you doubt five years ago that you would now argue for?
Peter (39:39.866) Yeah. I would say okay, let me backtrack on this one. I would say when I was working in in VC, I was very, very, very bullish on B2B SAS. I thought it was the best way to to play it. And obviously I was wrong, in terms of you know, when the AI came. But once I saw the AI aspect then I I I s shifted over and so it was at along that aspect.
Brian Bell (40:04.686) So I'll I'll ask the VC question then. So if you're a VC today, besides investing in sorcerer, what kind of companies would you invest in? What are you bullish on?
Peter (40:13.948) I think there's a lot of industries that are have a huge opportunity that are fundamentally to the fundamental to the economy. And so I think stuff like mining I think is not very popular, but autonomous mining is something that's very interesting of an idea that I think is cool. We saw Adams recently raise money from A sixteen Z. but I just think it generally as a concept I think is is quite exciting.
Brian Bell (40:42.252) Yeah. I think I think to build on your point is we are starting to not have scarcity around code and and software. We're starting to have scarcity around other parts of the economy. And so we're, you know, we're now seeking new frontiers to automate, right? Like mining, like manufacturing, like the you know, the the marketplace that you're describing, with import export stuff with with supplies and goods. And so, you know, what What's limiting used to be software. It used to be hard to go build software and you know, make it performant, scalable, extensible, all that stuff. But now that's getting really easy. I mean, I just vibe coded my website over the last week. It was really easy. it has lots of different modules on it, does a lot of different things, right? so I can just imagine that any you know, anybody who is sufficiently skilled can can really crush it on software development. And yeah, and so VCs now are we're we're seeking, you know, I I I personally just in the last couple of years have gone deeper into the stack, deeper into exotic, more exotic, unique new things like what you guys are building. So what what's a popular idea in in AI procurement that you think is wrong?
Peter (42:00.944) I think that people think that the labs will continue to make so much money and I think what we're seeing is that like what there's gonna be mythos type level open sourced type of projects and so I think that the labs are not gonna have as much defensibility as people think and I think that the margins are gonna collapse significantly. as people are it's gonna be like a race to the bottom. We're already seeing this already. But I mean if if there's such a huge difference between you know Claude or one of the Chinese models, if it's maybe two percent worse, but I mean it's eighty percent cheaper, then you don't really need that two percent. And I think what's happening now is the models are strong that that that difference of the one two percent doesn't doesn't really make as much of s might make much of sense. So that's what I would say.
Brian Bell (43:06.936) What's the best piece of advice you ever got?
Peter (43:11.245) This one's very, very basic, but customer discovery. When I started my first com company, like, well, I think and this was back when I did the pre-processing, I'm like, I think this is what people want. Like, yeah, I could talk to it, but it's okay. I don't really need to spend my time doing that. And it's just such a basic thing. Do the customer discovery, talk to people, see if it's a problem. And I was too overconfident. And then what happened? Boom. I d we we didn't have as much traction and then we ended up pivoting with toward from from the pain processing to doing the more CRM system and stuff like that. So that's what I would I would say.
Brian Bell (43:48.738) Al to end the inverse of that, what's the worst piece of advice that you hear smart people repeat?
Peter (44:04.716) I think there's sometimes people ha have aspects in terms of like patience is a virtue and kind of along those lines and I think at least my perspective is especially in the startup aspect is you gotta be aggressive and go what you want wanna do. And so I think especially right now, in the the time that we're in, I think now is the the time to actually be very aggressive and so that's what I would I would say.
Brian Bell (44:31.874) Yeah. What's a failure that you are grateful for now but hated at the time?
Peter (44:40.363) I think that initial the business that I was doing with the payment processing for the for the nonprofits. and then I obviously it ended up being okay where I I pivoted the business to doing like CRM and other stuff. I think that was multiple things that I I could have done better, right? I could have done better on the customer discovery. I could have done a lot of things in terms of the sales side too, right? Because I I come from more of a technical background. that I wasn't really as used to doing sales. And so with me doing calls and cold calls and it was just a lot of rejection that I didn't do as good of a job of just continuous on. And now today, obviously you know, more experienced than when I when I did my first like, you know, real business back in college. so that's what I I was grateful though for the experience because I really learned a lot and a lot of mistakes you know hopefully not not gonna make and be able to to approve upon for our our current business.
Brian Bell (45:40.792) Peter, I really enjoyed the conversation. where can folks find you online and find out more about Sorcerer?
Peter (45:45.545) Yes, our website is sorcererai dot com and my name's Peter Zitali and I'm also active on on LinkedIn. But thank you again so much for the time.
Brian Bell (45:54.947) Yeah, appreciate it, Peter.
