Arjun Bhargava, CEO of Rye: Building Checkout for AI + Amazon & on OpenClaw 🦞 and Agentic Commerce

Specifically, like an agent like Cloudblog and Openplog, I think it's more just kind of experimenting when we were building out the Amazon skill for Openplog. I had started out to my own Openplog and my Openplog has actually Amazon integration on my Gmail channel and my channel. I realized that I don't really really want to read things on one day and it ordered me from Amazon like a case of soda to come to my house around the time I was going to have meetings. I thought that was really really cool. Welcome to the retail jendic podcast where we explore the intersection of retail e-commerce and AI agents which we call retail gennig. Our mission is simple. We are shining a bright light on the new agenteic AI shopping mega-trend which will disrupt the trillion dollar digital retail market. The stakes are high, Google, OpenAI, Microsoft, Herplexity, Amazon, Apple, Meta and Anthropic are all fighting to be the consumer AI shopping agent winner. There are going to be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Could you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents? These existential questions are the tip of the iceberg on this podcast we're talking to industry experts, analysts and influencers to get answers to these questions, identify trends and best practices. This is retail gentic, your podcast partner in navigating our AI shopping agent future and now here's your host, Scott Wing. Welcome to this episode of retail gentic. On today's podcast I have an interview with Arjun Bargava, co-founder and CEO of RI. You can find RI at RYE.com. Here at Retail Gentic headquarters we have two ways that we find guest. Number one is what I call the thematic approach. There will be a topic related to gentic commerce. We find interesting and we will go research that and find a guest to help us learn more. Long time listeners viewers will realize that some thematic areas that we always find interesting are payments and how that relates to agentic commerce and also the intersection of ads and agentic commerce as a couple of our last podcasts have covered. Then there's inbound. We get unfortunately way more inbound requests to be on the pod than we can satisfy and where most of them are out of the bounds of what we talk about anyway. Occasionally we will find an awesome guest to that mechanism. This was actually a third way. A lot of you probably remember the episode I did with Ryan Ead where he helped us understand the open claw world and how that ties into agentic commerce. On there he mentioned to be able to check out something on Amazon. He had to implement the RYE skill. I think it's been a while but I think I said tell me more about a RYE skill and he said it's this company that has figured out how to check out on Amazon and immediately my spidey since tingled and I thought I don't know what this is but I must research this and find out how to possibly get them on the podcast. To make a long story short this is what I call the serendipitous approach. I just outreach and argent answered relatively quickly and I was able to give him on the podcast for you today. This is a really wide-ranging conversation. This is what's fun about founders in this space is they they're packing in like what I think of as like 20 years of founder experience in a short amount of time. We talk about how they found it RYE, Arjun's background. His co-founders got a very interesting brand that would talk about how they got the domain name. If you're not familiar with domain names finding a three-letter one-word domain name is exceedingly complicated hard to do. How they have single-handedly cracked the code on Amazon. As you know long-time readers will know Amazon is not the most friendly of agent at commerce folks I'll say. So they've got that feather in their cap. How they discovered the open-clock community and many many other interesting tidbits of information. So this was definitely a serendipitous diamond in the rough that we discovered for you are listeners and viewers. I hope you enjoy this conversation as much as I did and now here is my interview with Arjun. Welcome to this episode of Retail Gentick. I am joined by Arjun Bargava. He is the co-founder and CEO of RYE. Hey Scott. Good to meet you. Arjun, welcome to the show. Thanks. So I recently had you haven't seen it yet so you're going to have to believe me but I had a podcast with a friend of mine who our number one request is OpenClaw information and how that's impacting agent at commerce. So I had my good friend Ryan Ead come on and he walked us through in the last episode what's going on with OpenClaw and he was walking through some of the skills that he had that he found indispensable for making OpenClaw work across the internet and he said RYE and I was like spell that for me and he did and I was like that's a company I haven't heard of an agent at commerce. I must find who this is and reach out to them and get them on the podcast and you you kindly responded very quickly which I appreciate. So thanks for being a good sport and hopping on the podcast site unseen. Yeah of course I'm a big big fan of the podcast and the acts I did to be on. Let's start with your background. So is this the first company you started or are you a serial entrepreneur? What's the origin backstory? Yeah yeah good question. So first company that I've personally founded. My background is actually in software engineering. I started my career working in Fintech at Goldman Sachs as a software engineer there. I spent a couple years at Reddit, you grew up with an engineering team there and then did a few different startups that weren't venture scale but a little bit smaller had an hitch to do something bigger and that's when we started RYE actually back in 2021. So RYE is co-founded by myself. My co-founder is Justin Conn. Justin previously founded Twitch so let's say Amazon and we had an edge to really kind of solve e-commerce or go down deep in e-commerce. Because we didn't have a ton of prep I didn't research this but Twitch was Justin TV before Twitch wasn't. That's right. He's the Justin. He's the Justin. Justin's on TV what a camera on his head and then he was recording every day. Yeah everyone thought it was crazy but he was way ahead of the game. I think it's your head of the game if you started RYE in 2021 because in my mind my world starts in 2022 with the advent of late 22 with the advent of Chatchapet 2. Did you guys pivot to this or you guys saw the vision that there'd be a need for this back before Chatchapet 2? It was actually the initial thesis of RYE was to build Strike 3 e-commerce. An API for startups to sell any product online without having to deal with all the infrastructure of the e-commerce without having to deal with brand deals, payments and all those things. The initial thing that we saw back then was at that time called contextual commerce. This idea that shoppers are moving up the funnel, ways from going directly to say Amazon or Walmart or to the supply side and finding that context of a product directly on say a live video streaming app or from a theater if let's or from a friend on Discord. Justin because of his Twitch background he also has a fund where he invests in companies. If he was being pitched basically a new live streaming video application basically every week and all of them were amazing ideas but the kind of missing piece there was infrastructure to actually do the order in PhD. The actual e-commerce layer there and so we said hey it would be amazing if it could build a really simple to use API at that time. So yes it was before Chatchapet 2 before you know a gentick commerce was a thing but it kind of distilled to the same thing which is contextual commerce. Yeah that's interesting because I'm a big fan of what not and I'm not a huge fan of TikTok but I watched TikTok shops very closely and they've both done a very good job of marrying commerce influencers, entertainment and and all the things and you know they have that same situation. It's just one's kind of like collectibles even though I saw an article yesterday at what not getting into food I don't really know and then and then you know TikTok obviously is kind of like younger style beauty they dominate the beauty category so it makes sense now that Justin would connect those dots for sure and then so you said Goldman Sachs that's a good place to do Fintech if ever there was one. That in my mind says New York and then Reddit says Silicon Valley where y'all based out of. Rise is a fully remote company I grew in New York City but we have folks you know across the US. No subarriage NT is all the West Coast but I actually grew up in the Bay area in Cupertino so they have kind of lived between the two coasts. So you started the company in 2021 and then when did you when did you start leaning in on the on the agentic commerce side? Yeah, I'd say it was. really kind of started or focused on the geoptic cover side started in 2025 or 25. I think previously to that there was the idea that shopping was going to come to LLM, was a couple of jet-shoots etc. But I don't think the industry really realized how powerful a geoptic cover should be across all the different verticals here. One thing that we built or during our time since 2021 is we had a lot of integrations with Amazon to do Amazon Overrade, which Shopify to do, Shopify Overrade. But one thing that we realized is that we needed to greatly expand the kind of catalog of products that we could offer. And so in early 2025 we decided to kind of hunker down specifically work on building our geoptic solution to an image of the geoptic commerce theme where folks shopping or having their Asian shop on their BS aren't going to be looking at just products in Amazon or just products in Shopify. They're going to want to be able to buy any product across the internet. And so it took us a long time but we spent a lot of time there building for a geoptic commerce. So explain to I think I have a pretty good grasp of this, just in case I don't and just so the audience hears it. When someone asks you you're at a cocktail party and you know Uncle Larry asks Arjun what are you doing? What's the answer? Like how do you who's who's your customer and what kind of capabilities do you kind of talk about today? Before I answer that I think what's interesting is that to answer that question was really really hard more than a year ago. But now with help you know big with the geoptic commerce and I think how much people are understanding what's going to happen. As much as you're to explain. So you know now how I answer it is you know folks know that Chech of T is shopping today. They know that Chech of T mostly zealem's are going to have the ability for you to actually buy product directly within the application. Now to actually buy that product takes a lot of of work. There's a lot of of technology needed to make sure that it's the right price. It's the right product. You know choosing the variant that you need and then obviously actually placing the order that you place on Chech of T to the merchant to actually go for the fill. And so where I comes to the place is you know we take that layer of you place in order on Chech of T or any allegations that we work with. How do you get that order to the merchant to go for the fill? We hook up all the rails there to do that. And then so your customer you know kind of in that spirit you want to be striped for for the e-commerce layer your customer is agent builders. So I'm building a cool new video platform for you know some some vertical purpose fashion is the big popular on these days but that wouldn't be something I would do but I'm not really a fashionista if you can't tell. And I need you know so that I can focus on my value add which is like maybe the influencer you know recruitment and the platform for them you're going to provide the backend. Is that kind of like who your customer is primarily as the agent builders? Exactly right. You know agent builders are kind of a big segment there. AI applications you know folks that are maybe creating a user experience for folks that don't have their own agent but you know want to shop or or discover products building the kind of a professional here or slower applications like that. One interesting thing you know about rise again we've been doing this for a while previous to chat to you and a lot of project to commerce companies that are coming up now we actually do work in other segments or other verticals where that kind of fulfillment or orchestration later is needed and we actually use now a lot of the e-geptic commerce tech that we have to facilitate that so that's across things like B2B, procurements or social commerce you know categories like that as well. Same more about the fulfillment layer so is that where you're like actually getting into and I know enough about this to be dangerous but so feel free to correct me. So you know an order comes in and sometimes you can maybe just inject it into the commerce layer and it gets routed from there just like it you know you kind of like hack into where the site would place it but a lot of times you need to get into like an OMS warehouse management piece or sometimes people in my older marketplaces you'd almost have to go around that to kind of get it figured out is that is that are you kind of getting down into those layers or it's still pretty high level. Exactly it's actually kind of sweet of all of the bugs so we kind of look at it as a kind of a waterfall method here where the top level is if we have direct API access to the OMS layer or to the ERP layer we just face that order because that's been around for the last 20 plus years and we do that with quite a few retailers, suppliers, merchants for merchants that we don't have that relationship with that haven't you know installed with Rye or haven't done that because infrastructure worked there we do have an agent that can go take the order intent that was placed on the age layer above us take the payment information etc and actually go to the merchant and do it check out as if it was the shopper that was doing it check out and then kind of our final layer here is you know it's really important that the order goes through so we have we do have a managed fulfillment team that actually makes sure that orders a place correctly you know that that the very was chosen correctly it could place the order if necessary as well. One one funny story is when Ryan said I'm using this skill you know the context was an open claw skill called Rye and I was like so we had a little who's on first and he's like no the companies called Rye I was like okay and I went to try to find you and I couldn't find the URL at first because I thought certainly it's not rye.com and then it was because and the reason I didn't go there to begin with is three letter domain names are exceedingly rare so what's the story on the name and how you get that domain name? It was a story so to tell the story backing up a little bit around kind of a raising and in the early days of Rye so we were we were funded by interesting heartwits and we raised our seed route in 2022. Prudest to raising the seed route one of our kind of advisors and close friends in the business he actually owned the domain name called operator.com which is now owned by OpenAI for their operator application and so everything that we were doing at rye was under operator.com when we went into the cycle of fundraising you have due diligence etc. Ad reason had kind of an issue with us borrowing operator.com so they told us to go find a new domain name and so we actually introduced us to domain broker who gave us you know a bunch of different domains that were open to get a catalog our requirements were a one word.com name which you know obviously is rare but we were adamant about apiore.com domain out of his list there was one three letter dot com domain which was rye.com and as soon as I saw that I immediately was like we needed to call our company rye.com the problem was that I think because you like rye bread or like what what was the affinity? I do like rye bread but I don't know what it was it just clicked in my brain where you know for the next decade while building this thing they called rye just seemed really cool so you know it's just I fell in love with it there was a bunch of other interesting domain names or I honestly can't even I can't remember what someone thought it was weird but you only had eyes for rye exactly the the budget though or what it was listed that was you know a couple million I believe it was two million dollars. There was a crazy price on this it was a crazy price so we asked our neighbor up to see if you can reduce that price or get a contact seller he did and he brought it down pretty drastically where it was kind of no way for us to I like it it's easier to remember and it's easier to find now that I know it it is easier to find so it's rye.com for those of you listening so congrats on that that's a great domain name. Let's talk about you kind of introduce the topic so I'm going to kind of peel the onion on it so fun raising wise you've done a seed round from a 16z did you guys announce kind of the size of that at all or that's that's super secret. No yeah we did it it was a total of 16 million rye is from entry set and a few strategic investors as well as a few Joe Vesher's. That's a heck of a seed round that's a very very much Silicon Valley seed round right there. And can you say who the strategics are or they're they're not not announced at this point. Yeah we have a handful of folks that are deep in the e-commerce weeds one of our other investors is Felicia's who has been immensely helpful and a few other angel investors that are deep in the commerce we actually did approach release with forums back in 222 and yeah a handful of great green advisors that's a huge line. Was it hard pitching this idea in 2022? I guess at that point you were kind of like you weren't like riding on the new wave of a gentle commerce you were kind of in the yeah exactly yeah I think a lot of investors realized that contextual commerce was going to be a thing but where other so that was official. And having Justin as a co-founders always always helpful. Always a cool No offense to you. You're very smart too, but he has a brand. He does. The way I found you was through this open-cloth community. Were you aware that the open-cloth community? I assume Ryan, he was finding all this out on the, I don't know which of the open-cloth boards, but he found a lot of this, not directly from you guys, but through the community. Are you aware there's a vibrant open-cloth community? They're basically agent builders, just another flavor of agent builder. Their platform just happens to be the claw. Were you aware that they were out there kind of building on Rye or were you surprised by that? Our team is actually a huge open-cloth enthusiast. We were there when it was called the Cloud Vibe and you know when this really took off, we saw how powerful it was, you know, all the different news cases that people were focusing across socials. One thing that I wasn't saying was the commerce layer, you know, people talking about, I want my bot to buy something. So I believe actually the day that a clawed by re-named themselves, the open-cloth, we actually launched the skill for Rye to allow you know, open-cloth users to buy or sell anything from Amazon. So we launched that and that's a tool that Rye is using. Ah, okay. I kind of incredibly assumed some hobbyist kind of built it around your API or something, but you guys leaned in and you said we want to build our own tool. Yeah, that's right. Yeah. If we have time, I want to ask you how you use claw internally because I think this is endlessly fascinating, but we'll save that for the post-mortem here. Okay, so the reason Rye and you know, researched and found you guys was I gave him like what I thought were three hard task task number one was look for something like a TV that would have local inventory and then find it available at all the local best buys because local inventory is a hard problem to solve. And so that was number one. The number two was go buy something on Amazon and then number three was use vision claw. Have you heard of vision claw? I have actually. So some guy has hacked Rayban classes to talk to open-cloth and then he hacked it into you can use it for shopping. So you can hold up any object and say vision claw look for this online and then you it will talk back to you and you can order it to your glasses. That's amazing. And it will show you the product in the little window. Yeah, and I thought it was AI generated. Spoiler alert, he knocked down all three flawlessly, but he couldn't have done number two and then even number three, once he was connected to Rye, he could he could shop on Amazon using his classes, which is pretty wild if you think about it. And the reason why if listeners haven't and viewers haven't put this together, Amazon blocks most bot traffic. So that's why Rye was the unlock. So explain how you guys are able to solve that if you're comfortable sharing, you know, I don't want any proprietary information. But the extent you you know, you mentioned early on you were really tight with Shopify and Amazon. And imagine there's, you know, there's some kind of a situation there that allows you some form of ability to kind of help people, you know, search and check out there. So share more about that if you can. The answer is and I think as interesting as a lot of the highlights, you know, we again, we've been doing this that's twig-twig-two. Actually twig-twig-two, we pitch kind of this vision to Amazon and their teams at that time they were actually looking to API-ifying their ordinary solution. And they saw Rye as kind of a great thing there. So we have the ability to do that ordering through the Ramzon, the APIs API. The very thing about this is that it's not some patient or a bot going in and ordering products. It is a direct API access. And one caveat I'm sure Ryan has sprung to this is the way that we can place orders is through our account or my account. One unlocked that we're working on and I'm sure Amazon is working on it and folks need is the ability to bring your own Amazon account into your ordering. That's coming up soon. So you guys take all these orders, you filter them into your account, it gets placed on your prime account and then you your through the API, you're setting the ability to ship to appropriately for each order. And then on the back end you're decoding whatever you need to do to do the fulfillment or you know you can map the order. You have a unique identifier to be able to kind of unpack all this stuff going through one account. You pack it and unpack it kind of thing. Exactly right. And we actually will forward order confirmations, tracking information, all post order items to the address on the order. So folks get that it's too bad you don't get loyalty points for that account. I'm sure I'm sure that the GM the sales to that account must be up into the right. That's nice. That's a high class problem. Amazon thinks you're like the best buyer ever. We love we love right. They do love us and then I think they're also trying to understand obviously the chapter of the college streets well. So yeah we try to get partners with that. And show them what we know. They have this I've been studying Amazon for 20 years. They have this internal number and it's called a sugar rating per buyer. And it's and that's how they know when you call in for return they have serial return abusers and they have very low sugar and there's other things you can do to be low sugar. Sales volume is like a big a high sugar. You may be I'm imagining there's like some sugar scale and you guys are like way off the chart on the you're like you know a thousand percent sugar or whatever the metric is compared to everybody else. Really interesting. I kind of assume that's that's exactly how they would rank customers. As you get further up there it's very Amazon you know anything they measure everything and then create incentives on both sides and you know so there are certain buyers my other podcast host Jason Goldberg I think he's got you know you're probably right above him on sugar he orders so much from Amazon and and you know he almost can like pick up the phone and the answer instantly like I think it actually you know it actually bumps you up the cube based on your sugar rating like all you know imagine what you could technologically do with a rating like that is like a you know a customer favorability number and they do that kind of thing. So I'm imagining you know someone on your side calls they get a quick answer. And I bet the return rate is actually lower because people do so much research when they do a gentle commerce. You know I have a I have no data to support this it's just kind of a thesis and I've been looking for data. I imagine this the return rate is actually insanely low which is good. Our return rate going into this quarter is is sub one percent for her framzon and I think that's the thesis is that you know folks up the funnel have more contacts there they're getting more research around the product and so it's you know it's a much better buying decision than you know potentially just randomly shopping or basic orders. It makes sense not to know the backstory it makes sense because Amazon made a big push based on the success of what not and some of these other and in China like the video platforms are the big almost everything's done through video and TikTok shops I know Amazon's had several you know attempts at doing their own video stuff but they're also big on they have a whole influencer marketplace type thing too that you guys would be a perfect add-on on the back into that so it makes makes sense now that I kind of understand the backstory. So you mentioned earlier you've gone through a phase to kind of like do a much bigger expansion. What so you mentioned Shopify and Amazon what's that mean is it like Salesforce Commerce Club and Big Commerce and that type of thing or more you know let's say I come to you and I say I'm doing auto parts agent if you already get all the auto part you know I want US auto parts I know a lot about this you know US auto parts auto zone you know this that the other or would you have to go maybe train your agents on those for a while because they're retailers you haven't seen. Great question that interesting we've heard of the auto part when we have we have customer that is integrating out I believe they're actually wide and that is the marketplace for auto parts and you know the soldating them and there's surprisingly a lot of auto part suppliers on shopwise they're actually like in a lot of shopwise stores but to answer your question so we kind of approach it as two different paths so where we can we'll do a direct invasion so on our roadmap we don't have these live today but on a roadmap we had a plan of the nation with Salesforce Commerce Adobe Commerce Big Commerce and commerce including kind of a more custom ERG integration that allows to you know when suppliers want to directly integrate with ride allow us to start selling direct to that but the second approach is our injective approach and that is our our ride aging as bullet that is able to go to a supplier that has obviously a public presence I guess check out method and actually place that order directly at the website and depending on where or what our customer or developer is looking for they can use MUBs cut tools to place orders and I have a very vibrant set of audience members that are in the payments world and they they would always they're screaming right now Scott ask the question how is he filling out the credit card isn't that a violation of PCI security so I I can tell you've already thought of this but I will ask the So I can I think I know the answer, but I want to hear you say it. Yeah, you know payments have been a phase For a very like time up until like say this year so historically We ride actually act as much a record So we have our own striper counts. Well except payments on behalf of our Developers which call our customers our developers customers actual shopper's and then we will take The actual and supplier the merchant was a corporate category or payment method that we have internally Obviously reconciliation there is a mess as you could probably imagine or think about there There's two different transactions there. It's not economical and it of course is a lot of friction between those things And when talking about return to refunds, we have to do kind of a two-step process to make sure that that all works Really kind of middle last year So so that that your merchant a record and you're probably creating a unique stripe account You're probably using a virtual credit card like mechanism. I'm okay Virtual credit card because you're not gonna take a your stripe your company card Which has like you know 300k of and you're and you're not shooting that over again the payment people would be very mad at me Yeah, that's a question. Yeah, very good call. Yeah, that's that's really really important We have card rolls on our virtual cards to make sure that it's tied to the exact single use X dollars Bam yeah, exactly, but but still you know not far from I to look perfect Really though the last year you know that in the I see In mesh charts organization technology as well, you know, there's an unlock now where agents and You know rise is gonna feel like across that we do Can take the actual payment information from a shopper tokenize that payment information and that's that to the merchant for the merchant to To actually authorize the charge We are integrating with an ancill of vendors today to to authorize that There are still definitely limitations the breasts of payment methods that we can support So you know, it's not fully across all of rice capabilities or you bet, but we are now You know, we are now tokenizing Piedemann that is saying at placing on a merchant up strawberry. Yeah, and we've had master card on the program And I'm intimately aware that visa and am X all have you know, they're really excited about this world I imagine they've reached out to you and maybe you're you know if you're comfortable saying Do you doesn't make sense for you to kind of like go into that layer? You know, they they all have names where I forget they all have their own protocols to This is this is where it gets kind of past my payment knowledge Doesn't make sense to do that or the merchant is better to send it to the merchant than they can kind of like then route it to whatever from there We don't want to Get his that layer directly Yeah, there are great great vendors Parties that that are trying to tackle that problem both from Top of the funnel side before it even hits rye to be able to tokenize for carrots all the way down the funnel to the merchant side Make sure that that credit card is the taxable shoppers credit card at the top of the funnel So we're part of with this at both closer another set of people are what I would call this is really interesting I don't see this with ACP but there's a Vibrant set of UCP enthusiasts. I'll call them UCP Maximist You probably have met the the crypto Maximist to the Bitcoin Maximist I'm now coining the UCP Maximist and they're gonna look at what you're doing and they're gonna say UCP puts this dude out of business. There's no need for them. So so what's your I know that's not true But I can like they're screaming when the payment guys are screaming. They were screaming too What's your what's your answer for the UCP Maximist? I think for both UCP and ACP You know, it's Really really interesting particles and ski mugs. I know a lot of thoughts obviously there's been a lot that's happened over the last six months For both of them. I think it's fundamentally the way that we position ourselves We think about ourselves is that if Either you or you do truly take off and the vast majority of versions Decide to integrate or have an intupera acp Kind of connection. We would be on the opposite side of that. So you know It's that is the world then you know, we would have a layer where we would be able to plug it to any ACP Connected region similar to how we work with Shopify versions today You know, we are not just placing Shopify. We work on top of Shopify and Shopify is amazing at Cataloguing and kind of Aggrating all of these great merchants and you know, we can now have access to that directly So I think that would be the future with both acp and UCP I think the maybe the the more hot take here is that I think kind of unified e-commerce protocols Are not really a new thing. I think there's there's always times where folks believe that every merchant is going to go and install One protocol or One integration and I think that really effort does work. So I think we're kind of seeing that right now where retailers merchant and then I would love to hear your opinions We'll stop. I know I know you talk to you some merchants, but uh, I My my thesis or God is that a lot of merchants are hesitant to do the work here now Of course Shopify is really bad at house and I'm sure sales force commerce and see other companies are as well, but Yeah, it's still to be seeing if if the student take us. I'm not a UCP Maximus because I've been in e-commerce long enough to know there's always You know, there's always a reason a cfo will say no and you know So you know with this head and tail problem. So there's a bunch of head people. They're gonna say no And that's gonna be like my head people. I mean like top 50 retailers. Here's an example amazon You've already solved that so good for you So so amazon is not announced. They're gonna adopt any of these things and if they do it's gonna be in an amazon way meaning it's gonna come with a lot of Yeah, it'll probably be their own flavor of it or something And there'll be some other big companies that don't adopt it as well and then there's this long tail and like the auto parts is a good example, right? So is cheg.com going to adopt one of these protocols? Probably not and they're on their own e-commerce platform. It's fine for their needs. Maybe it's like even I don't know what they're on but you know There's no there's no way to poke into that versus there's always gonna be a need for the agent form filling thing I think if you truly if you truly want to deliver to your customer coverage of anything And not just like the fat middle, you know if you if you you know and a lot of the best e-commerce happens in the tail You know Shopify is a prime example of that. They're basically like an encapsulation of the tail They're starting to get a little bit of head head coverage, but it's not it's not that much of what they do So so I think I think you'll have a business for a long time You can you can sleep well tonight. You see he's not gonna play you out of business for shit Seems like it wasn't buying you that much anyway Yeah, so where do you see? You know where do you guys want to take the company? Can you tell us kind of like how big you are now? You're you're kind of like you know kind of five years into it, which is you know Congratulations most companies don't make it five years especially and this something changing as fast as what we're doing So congrats on that Have you guys done An A yet or just on riding seed at this point? Yeah, that's right We've actually stayed in Kredelwellian since 20 or 20 years to raise And you know, why do you see continue to do so a lot of the events of AI is help Not to their shop. He likes to talk you talk about that, but You know, we we are not possible, but you know, we we're we're good right now from our RC clock raising we're a team of about 12 mostly engineering actually As you know a lot of the work here and it's just really really tough to imagine about self So we most like a well raise our our series a in about a year or so But right now we're just letting it build I see a lot of founders that raise a big seed and they say it's gonna last four years and they burned through it in year So congrats on not falling into that trap and then You know going that long is just amazing and you know, it's interesting because we did come like in 21 it was you know Scale the company as fast as possible and now it's definitely revenue per headcount So you're you're kind of on the right side of that staying lean thing That that used to be you know a sign of progress and kind of Silicon Valley land You know a big raise and a big headcount and now it's like the opposite way around so congrats on that So that's two pitfalls you've successfully avoided The What do you see as the future of the company? Is this? You know you guys Like are you you you're obviously feeling the energy we are around agent to commerce and Do you do you think a lot of these you're you talked more agent builders than I do Are you or some of them starting to get traction or they still it's kind of like this we're in this like you know Primordial phase where there's just like a soup of startups doing all kinds of crazy stuff They starting to see some like everyone talks about daydream That's like the one this you know seems to be getting some some pull away You don't have to say names, but are what kind of interesting things are you seeing out there in the landscape of agent builders? You're talking to that I'm not definitely tip to answer the question. You know it is there Kind of that it's like shipboy have name definitely that's that's what we see You know, I think you know we've worked in a lot of different industries, like I said, I would press like, for example, or we're geistic, for example, social commerce. And we were kind of a player that mixed up lots of different commerce, you know, infrastructure companies to do that. Really the last year with agent builders and AI platform is trying to effectively eat a lot of those kind of older segments. I think we're saying kind of a massive push where where e-commerce was transacting previously, it's now going to be weight-war, fish-in-weight, faster, really better with agents, with AI platforms that are going to be verticalized and can do, because e-commerce pieces, much, much faster. And so that's what gets me really excited, I think gets a company really excited, is that it feels like we're just like earlier, early earnings were starting to see that growth. And as AI gets better and better at faster faster, the companies that we're doing hundreds of millions, $2 billion, June, and like the procurement space, there's going to be an agent or a company that could do that way more effectively. And if we're part of that, that's pretty really cool. Yeah, kind of going through e-commerce 1.0, everyone's slept on B2B. Like I remember when Amazon Business came out and no one thought they'd be successful. And now it's like $30 billion or something. It's like so big. And I think that's going to be your bigger opportunity as the B2B, because you know, imagine just like, every vertical has this procurement thing that gets into like very specific stuff. Like I know a company here local that helps dentists buy things. And you know, these machines that dentists buy, it's like make or break for that individual dentist office. But it's like super complicated and navigate. And they all come with their design to create, it's like a confuse opily. They're all designed to be so, it's like buying a mattress you can never price up. It's all designed like that. And just like having an agent that advocates for them and then like helps them navigate these complex procurement decisions. And every vertical is going to be, you know, a huge opportunity. And then having you guys like, so the agent builder can focus on the super vertical subject matter expertise and you guys handle the back end. I think that's really, I'm going to predict that that's going to be bigger than the consumer side for you guys at this. Yeah, I don't know when. The hardest thing predicting in this space is the timing. Because in the traditional world, like in e-commerce, it took 15 years for B2B to wake up. And then when they did, they kind of got going pretty quick. But here 15 years means three years. You're already five years in. So could show up any day. Yeah, that's interesting. In B2B, a lot of times they get involved in like, and I don't know too much about this. So there's like this whole CPQ, which I think is customized price quote. Is that something that you think will live up in the procurement side? Or do you have some of those builders like you're kind of wanting you to get involved in that? Yeah, we wouldn't be involved. We do work with, or kind of an early innings that we're working with a few folks that are working into that space. You said, kind of touch on this. That is a really difficult problem. And they just saw that. Or that's really going to be their value, very proposition, you know, really pay your eye to that problem. So it's over on the agent builder side. You're not going to-- Yeah. And I will make sure that it actually gets fulfilled in order to start through. So you're kind of check out down almost always. Look at it. It's good to know what lane you're in. A lot of another founders problem is they get too diffuse in their vision. And they try to do everything. And then they paint it better and out. And it's a problem. Where do you see-- you're obviously marinating in this world like I am? I don't know if you saw my predictions post or anything. I kind of get pretty specific in what's going on at chat to PT and stuff. Where do you see the agentic commerce going in the next whatever time frame you're comfortable with? I feel like this space is moving so fast that the type frame post six months, it's just going to be possible to predict. I do think that there's obviously so much capital being pushed into, which are the decopers into building up infrastructure. I don't know what that means. That's going to be the next six months of just building. And it's still asking that question. Where's the volume? Has the contrast issue from E.C.E. Quattro into injectic commerce? But you also kind of touch on this. I do think that there could be a slip of a switch. And there's going to be a ton of volume. Once one of these providers figures out how to eat one existing kind of incumbent. And I think that's really exciting. I think consumer retail probably is not going to change too much in the next six months. But going back to the B2B side, I can see where there would be something like the order of 100 million annual SGMV moving to an agetics system, versions a normal kind of user in someone calling on the film system. So that's really exciting. That's a good thing to keep a track of. 100 million G and V from somebody on agent. All right. Let's go. I like that. I may borrow that. You heard it here first. I'll always attribute it to you. The one thing I wanted to circle back on is-- and if you're-- I realize this may feel proprietary, so feel free not to answer this. But you mentioned you've got a lot of open-cloth, enthusiast. I'm endlessly fascinating about how people are using this internally too. On one spectrum, you've got like Jason Calconis. I don't know if you follow his stuff. He mentions it less on all end and more on twists. But he's basically created like-- he's got a Mac Mini loaded with a claw, and it has a different persona. And he's got like six of them basically in a human persona type thing. So that's one of the spectrum. And another one is more like what Ryan Eats used it for is more of like an EA or a chief of staff type thing. How are you guys using it again to extend your comfortable sharing? Yeah. No. So I think it's-- it's internally, specifically like an agent like Cloudbot or OpenClaw. I think it's more just kind of experimenting and people doing kind of random things. Mostly on the EA kind of side. I'll give you my example of how I used it, or what got me really excited when we were building out the Amazon skill for OpenClaw. I had set it up to my own OpenClaw. And my OpenClaw has my-- actually, Amazon in addition to my email and my teller. And it actually realized that I don't get really, really long day of meetings on one day. And it ordered me from Amazon, like a case of soda, to come to my house around the time that I was going to have meetings. It was Poppy so-- and I don't really like Poppy so does. So it was kind of the wrong product. I thought that was really, really cool that-- so you're good, totally unhinged. So you're just like Leonard. Exactly. Yeah, fun. I gave it a car realized how much you could spend, and just general kind of things that I like. And chose that product, tell me everything. But it's a really cool experience. Is it you like some terrible flavor of popping? It's different. It actually did. Funny. Cool. Well, Arjun, thanks for being such a good sport and hopping on under such short notice. Congrats on the company. And all you guys have done-- yeah, I definitely want to get you back in like a year to see how it's going and see what other cool innovations you've done. Thanks for coming on. Really appreciate it. Thanks for the invite, Scott. Really enjoyed the conversation. Follow rate and share to stay up to date with Retail Ginic. Your guide to the future of AI Shubbing AJ.