Getting Started with Agentic Payments on Crossmint

Transcript

In the earliest innings, [clears throat] everyone sort of kind of assumed, yeah, at some point agents will probably pay for things. But as soon as Muse and Instinct came out, like really [clears throat] backtoback and and even in those like that first week, one of the first things that Muse suggested, gave you three things. It was like, "Let me check your email, let me do X, let me buy you something." It was like, "Oh my god, here's the inflection point." And we went from like agents will probably buy something to one day to oh of course agents are going to shop for us. >> Welcome to the Devon [music] Help podcast where founders, hackers, and anyone in between showcase their products built on [music] the bleeding edge of tech GMG. Welcome to the show, Alex. I'm ecstatic to have you on today, man.

>> Thank you for having me. It's um uh always so exciting to talk about what we're building here at CrossMint, and I just genuinely love telling more people about um what the future's going to look like. Absolutely. Absolutely. And yeah today's topic and what you guys are building um it's really a emerging um emerging space right now like everyone is really uh hot on like um what models are coming out every week but I think we're we're at a point where I think we have we have enough enough intelligence that people can start building uh building things on top of it and uh yeah uh payments payments are a big one but um >> it's funny you you put it that way actually was at a dinner with a bunch of AI folks last night and

>> yeah, >> um there were a lot of us kind of sitting at one part of the table saying, >> "Do we need these things to get that much smarter or are we still just like we still have to just mine deeper to figure out what are all the ways that we can put this incredible intelligence that we already have to the optimal work. So, um, yeah, I think that like the the tools that are already in place are incredibly capable and, um, we're we're ready to build some really cool things with it. >> Absolutely. Absolutely. And I I've seen my fair share of that as well. Um, but before we get into all that uh, great stuff, um, would you just like to give a introduction about yourself?

>> Absolutely. So, I've been in the tech world now for almost 15 years. I worked at a startup accelerator. uh I was um [clears throat] moved on to work at an IoT company building hardware and software uh to remotely man monitor telecom sites. And it was actually while I was at that company in 2016 that I read some of the um early work thinking about how Ethereum and smart contracts could be used to enable machine tomachine payments. And that's what kind of sent me down my my web 3 journey uh back then was the idea like I'd known about Bitcoin for a long time and was interested in it. But the idea that we could put make money more programmable

and let machines control money was one of the original things that got me interested in this space. I went on to work at chain analysis for almost three years. [clears throat] I then left chain analysis to start my own company. I built a protocol called demo um where we were putting machine identity on chain and had the dream of eventually getting into machine payments as well. Although we were solving a bunch of other problems around data management and data verification [clears throat] and after five years of building demo uh I left demo to come work at Crossman and then lead agent payments here at Crossman. So it's been a journey that um has taken me in and out of the web 3 world. It's taken me in and out of payments. uh it's taken me kind of like in and out of a few of these different areas, but

somehow I always find myself intersecting with how are how is money getting more and more programmable and how are you know non-human actors going to move money? >> Absolutely. And and yeah, that that's that's quite the amazing journey and yeah, I think you landed in in a great spot um as um I believe that this inter intersection of blockchains and and AI um I think is is a natural fit. Um I like I mean we haven't been able to get uh humans uh uh quite on board with with using wallets and and private keys but um all these things are kind of very well set up for uh AI to be able to understand and kind of uh work with. So uh it it is what it is. [laughter]

But uh yeah uh I I suppose um from that standpoint like u we know a lot of a lot of people um probably people like you and I and others watching today uh we we know a lot of Agentic uh products are heavily um developer oriented and um I think that's just like the near-term opportunity and and and it's what people are are paying for now. But um a lot of people don't realize that um agents will also eventually start to spend online and and um kind of do uh be um extensions of ourselves >> and um I guess I guess for the person watching today that doesn't kind of realize that um did you want to kind of

give a overview of um what your perspective is of that and um whether or not we should be giving agents our our credit card to to spend our cash? >> Yeah. So, I'll I'll talk a little bit about like maybe [clears throat] AI as a journey because I think that the moment that like took us from thinking about LLMs as a chatbot to really thinking about agents was of course OpenClaw late last year. Um, and I remember [clears throat] getting into OpenClaw and starting to build my own agent. And I'm a relatively tech forward person, but man, I was like, I'm not a dev. I I have no interest in the DevOps like [laughter]

and I started you know doing you know creating my own skills and then you start seeing other skills libraries and it was like it didn't take me very long for me to just say I would love for somebody to make me an agent with a thousand skills built into it that are already really good and just come along and give that to me and you know I I've used everything I used Hermes and that was like a step change but I [clears throat] think that obviously like the new wave of consumer assistance is like finally hitting that level where maybe it's this version, maybe it's like one more iteration, but like we're getting closer and closer to actually crossing the chasm and [clears throat] the um it's really funny. We've seen we you know Kier Crossman we've seen some

data many of the early open claw people um were [clears throat] you know iterating [snorts] on the edge and were playing it fast and loose and basically said sure I'll put my I'll give I'll give my open claw uh access to my credit card and I'll just save the card number someplace that I think is relatively secure. [laughter] Um there were quite a bit of there's quite a bit of credit card leakage in that world. um that is maybe fine. You know, if you want to take on that risk as an individual, that's okay. But like if you're trying to offer an agentic service to somebody else, there's a reason we have, you know, we have some compliance and regulations around how you should handle those card numbers.

So, um, you know, as the the in the earliest innings, [clears throat] everyone sort of kind of assumed, yeah, at some point agents will probably pay for things. But as soon as Muse and Instinct came out, [clears throat] like really back to back and and even in those like that first week, one of the first things that Muse suggested gave you three things. It was like, let me check your email, let me do X, let me buy you something. It was like, oh my god, here's the inflection point. And we went from like agents will probably buy something to one day to oh of course agents are going to shop for us. And uh the last like just kind of high level observation I'll give um just to get things you know started and get the ball rolling here is um I see a lot of people

talk about like what is the thing that agents are going to buy and there's a lot of like debating and arguing on Twitter. Is it going to be flights? Is it going to be groceries? Is it going to be this? Is it going to be that? The reality is I've talked to hundreds of people about this and people are all over the place. Like for me personally, I'm a Delta loyalist. I don't think booking flights is that hard or complicated. I have some nuances. I changed the way I want to buy on a few factors and like it's just easier for me to go book that myself. Like just easier. Um, I don't like I I where I have found agent commerce to be useful is on things that I periodically reorder but like don't make sense to put on a subscription. And so my agents like know

a list of things that I buy and I can say reorder the dish soap and the multivitamins. And they're just like, "Okay, cool. I know what you want. I'll go to Amazon and get it for you." And I think if you talk to people, they're going to give you a range of answers. There's no right answer to what agents will buy first. Um, and that of course is what opens up the opportunity for many many many many builders and for thousands of companies to to do this successfully. >> I I absolutely agree. Um, I think there there's an ocean of opportunity. It's just a matter of figuring out what what people want and and how they're going to go about using AI in their in their day-to-day lives. Like um developers are developers are one thing, right? we we we kind of have our uh have our like

daily habits of things we want to do and and how we like to use our agents. But um for the average consumer um th those habits are are going to be different, right? It's like maybe somebody just doesn't like standing in lines, right? And like they just want their agent to go buy them groceries from Instacart or something like that. >> Um or or yeah, it could be just the little things that you want to save time on. Um, yeah, that, yeah, it's going to be very interesting to to kind of see it play out. Um, but, >> uh, I know, uh, Crossment, uh, is kind of at the forefront, uh, trying to prepare prepare for that eventuality to to get, um, agents being able to buy things, uh, for us. Um, I guess, did you want to, um, kind of give an overview of

what um, Crossman is overall and then we can get into agent stuff after that? >> I'd love to. Um, so Crossbin's been around for about 5 years, and the company began as um a really easy and seamless way to run an NFT launch. [clears throat] One of the original innovations of the team was to make the um the bridge from fiat to NFT purchase a lot easier and faster. And so they were really the first team to create a headless checkout where a user could show up with a credit card or Apple Pay and purchase an NFT.

That [clears throat] eventually, you know, NFTs had their cycle and um not quite as prominent of a of a commerce category as they once were. Um but of course memecoins have become quite prominent. And then like the largest growth area of course for um really any kind of crypto activity today is in stable coins. And so over the last few years Crossman has become a stable coin company. We facilitate on-ramps and off-ramps to some of the largest companies in the world as well as wallets. So Moneygram, Robin Hood um and a bunch of other fintech companies.

[clears throat] About two years ago, actually coming up in November will be two years. Um the company quite a bit before I joined the company published its first foray into agentic commerce [clears throat] which was called the great onchain agentic toolkit go it was an SDK and the [clears throat] premise behind go actually still lives in our products today um which is how do you act how do you um give an agent a wallet in a controlled and a permissioned manner Crossmin as a wallet provider builds smart contract wallets. [clears throat] So we actually encode all of the policies that are relevant to

controlling those wallets on chain. Um and we you know use delegated signers. We actually have multiple signing layers. We have operational signers and then recovery signers. But basically it's pretty simple. If you want to give your agent access to your a wallet, provision a wallet to the person and give the agent a delegated signer. And in fact you might give it multiple signers. Maybe it has different budgets and it has different purposes for each signer and maybe you have different agents that each have their own signers. It's a very flexible construction and so we were able to put this out almost two years ago as [clears throat] a way for you to equip your um you know chat GBT with a wallet.

Since then uh the agent team has iterated and evolved quite a bit. Um we still support and we'll get into this in some demos and some some other activities but we still support wallets for agents. uh including on-ramps of course. So you get all that in a nice clean package if you need your agents to transact in stable coins. Uh and we've added cards to the mix. So we support some of the latest uh agent commerce protocols from Visa and Mastercard. And [clears throat] one of the really interesting things that we're seeing is virtually every payments company on Earth is now starting to unveil well they're planning to unveil some way for agents to access those funds. So, all of the big payments names that you can imagine are going to say, "You have an

agent. Here's how your agent can use your fill-in-theblank account, PayPal, Cash App, whatever." [clears throat] Um, the last part of what we've been building on the agent team is a checkout API because once you've given the agent access to some money, it still has trouble buying things. Um, in the world of checkouts, we can talk about this more in a bit, the world of like how your agent actually goes and executes a purchase with a merchant, uh, is not a very well paved pathway today. and it's one that is set up to evolve quite a bit over the next few years. >> Awesome. Yeah, that's that's all exciting stuff to me. Um especially uh you guys announced the um Asian commerce toolkit I believe yesterday. Yeah.

>> Um do you want to kind of uh tease what that is and then we can get get into the screen share demo? Yeah, I would say it's both a combination of the tech um itself and sort of the like the user journey. Um [clears throat] so the the the products themselves are this the the agent cards that we have, the agent wallets and the uh agent checkouts API. The uh major changes include one finally making all of that self-service. So um you know being a regulated financial institution and selling regulated products like on-ramp off-ramp most of what we do most of the business we do is not self-service you know we have to go

through KYB we have to spend a lot of time with our customers not quite as cumbersome as getting a bank account but we have just compliance obligations we have to fulfill um that's not the case uh with these products now there are some um risk checks and risk considerations uh especially because we sort of unique access to Visa intelligent commerce and Mastercard agent pay. Um so we [clears throat] have to be quite considerate about how we you know handle those things but at least we can help get developers to production without throwing up a gate in the way and that was really really important to us as we were bringing these products to market. Um so now we have like the full package of the SDK. We've made it all self-service and really crucially we've

put out not one but and not two but three sample apps and we've made it as easy to fork as possible and for somebody to get started and say I want to build a wallet. I want to build an agent that does XYZ or I want to build an agent wallet like here are all the different ways you can make that possible. And so you know that's what we're going to talk about here for the most part in the demo. Um showing all the different cool sample apps that we have. this is the way you got to make it easy for for developers to get up and going and and shipping something. Uh as there's lots lots of intricacies um in in kind of getting it getting this side of the equation set up like you think like >> high level this would be like once you

start like oh hold a minute [laughter] but uh yeah I I love to see this in action. Uh so uh let's uh yeah let's get into it. >> Let's hit it. All right, pulling up my screen share here. [clears throat] So, we're going to start with our um this is our like sample app. Um it's probably the most um uh it's the most flexible sample apps and I think it does a really nice job of showcasing a a handful of the flexibilities of the particular flexibilities that we have. So here we are looking at um uh and and the we should probably get this uh URL sent out because it is quite a cumbersome one.

[laughter] Um but we're looking at this sample app and right now it is set up to display what would uh an agent um iOS app look like. So here we have a mobile app um just to showcase hey like one of the really important things to us is everything we do is headless. Um, we are not bolting on like one of the most common things you see in the agent world today is many agents are using the Stripe link wallet as a means to accessing payment methods. That's a great product. Um, but the, you know, to to me it violates the cardinal rule of building a consumer product, which is you are now taking the consumer out of your app and you are bringing them to a third party app. Um, so what we've done

is we've built everything to be white labeled completely into your own product interface. So whether you're building um, you know, your own agent on a mobile app, a desktop app, uh, whether you're just creating an iMessage agent or even, you know, a WhatsApp agent, or making it a chat GPT plug-in, [clears throat] Crossman has you covered in terms of how this agent can uh, fit your brand and fit the user experience that you want to create. We have two separate apps. So the one you're seeing here is our cards demo [clears throat] and then we actually have a separate one uh that we can send out which is our machine to machine payments demo. So on the one hand we're saving a card in order to purchase

things on the web. And [clears throat] in this demo we're creating a wallet uh topping up that wallet in order to use x42 or other um other resources like that. You can see I haven't put any money in the wallet yet, but if I [clears throat] wanted to, we could we could take a quick look at that. Some of the other functions that we have. So, you might have noticed I have a wallet already saved here. Uh, in my case, this is my kind of daily spender, my my MX Sky Miles account. I [clears throat] could add multiple cards. Uh and one of the other really important things of course that you know even some of the folks like us like really on the edge of the technology are are getting used to is the idea that how do I control how much my agent can spend

and there are a variety of ways you can do that. Um [clears throat] so the card networks themselves are starting to build some of these controls in into these agent protocols. Uh and then we actually layer in in some cases additional controls via [clears throat] our SDK um where you can set a budget, you can set authorization gates, you can create additional controls for the user. That's part of what you do as the developer building on our platform. You get to decide like which of these functions you want to uh you want to use because of how we handle that authorization. It allows you to create a lot of really interesting and super flexible um basically like super flexible outcomes. So, you could pre-authorize an agent to purchase something uh and say, "Hey, I want you

to buy me this thing, but like wait for the price to drop uh from $60 to $50 and then buy it right away." And that's I think like one of the directions that people are going to go and like use agents um much more adaptively. I would say like, and this is anecdotal, but probably 90 95% of the agentic purchases that happen today are people who say, "I want this thing, buy it now." But we're starting to see an increasing share of I want this thing, watch the price. And I think the holy grail, maybe there's levels beyond this, I don't know. I haven't really thought too hard about it, but like the next layer beyond that would be, [clears throat] okay, agent, you have a budget. And I want you to figure out like which of

these things I need and when to order them. And I don't know, maybe one day agents will get credit. Like there's there's a lot of like weird weird things that might happen in the future. Let me take a quick pause and see if you have any questions so far. >> Uh I was Yeah, this is really awesome. Um love the demo. Love the um uh how slick it is. Uh I was going to mention um that that wait for the price to drop. Uh I I I've done this in the past. Uh I don't know if you remember when the PS5 and Xbox came out and you couldn't find it like in inventory. So I I personally made like a selenium script that would go and scrape the the different websites. So um basically this kind of

evolves into that and uh now that we can kind of um make it so that the agent can purchase it for us instead of just alerting us to go um to go check it out. Uh yeah, I think it's pretty crazy. Um and uh so um this was kind of the side of um equipping your your agent with a card, provisioning it for with spend. >> Um but when it comes to actually going and doing the the checkout part, the payment part, uh I'm really curious to kind of dig into how you guys uh solve for that because that's that's one of the bigger um bigger pain points um of like one one is like >> what can an agent buy? That's that's one part of it. And the second part is like

a lot of um uh checkouts checkout pages kind of block the block an AI from being able to enter details there. So yeah, keen to learn uh more on that. >> Yeah, it's a great question and we'll progress [clears throat] through a few of the other pages I have queued up here. So just as a quick orientation on the docs um we have a section for agent cards and this of course governs and explains kind of how our agents card product works [clears throat] how users can come and save cards. Um technically speaking there are some different types and different ways to save and use cards whether depending on whether or not that card can be tokenized by the card networks. [clears throat] This is a term that they have in terms of how the card gets saved. And basically just you know

I think the relevant point for developers is that we have secure ways to handle the card whether or not it can be tokenized and we can fall back to different pathways that matter [clears throat] when it comes to checkouts we just another note for developers out there cross [clears throat] its we don't handle the discovery layer so we're not you know we don't have a catalog we may eventually fold that into the platform there's a bunch of folks that offer catalogs out there but if you want your agent to have up-to-date information on what's in stock, what's on the store, there's [clears throat] other APIs you can go hit. We become relevant once the agent knows what it wants to buy. So, basically [clears throat] what you'll do is you would um actually submit, let's get into the guides here.

[clears throat] Um you actually submit a URL of like what you want the agent to buy and you can actually provide also some information um just some context about that. So that context is typically stuff like, you know, we have buyer profiles, so I might have a home address and a work address. Which one am I shipping it to? Do I care about the [clears throat] the shipping time? Do I um you know, if it's, let's say, we're ordering an item, I might give it the context on the size. So, I actually want to buy this pair of gloves. And uh I'm going to say uh standard shipping, use a card, XL. Now, as an agent builder, you could

store all of this information. Like, you could have user preferences saved. I could have told this agent ahead of time. Always use standard shipping. If you're ordering me gloves, Excel, just use a card. Um, I'm specifying card because we can also access other payment methods like shop pay. Um, save cards on file, whatever it might be. [clears throat] So, this is yet another one of our like demo apps that we're going to be releasing more publicly soon. We call it open commerce. You can actually plug it into your own coding agent um or just you know your your bot in general. I have it hooked up to Grockbot and it works amazing.

So [clears throat] what happens I've now given my agent an instruction to go buy this and it's calling the crossman APIs. Um this is a fun demo because you get to see the actual browser automation happen in person. Um the way things work today is that 99.99 whatever percent of the web is inaccessible via API. basically other than like travel most of the web is not accessible via API at all. Um so we are starting like our our site or our our tooling already does this. We check for one of the agent protocols UCP, ACP, etc. But we didn't find that here. So we're going to move on. Now we have to handle this checkout with an API sorry with a browser

automation. We have built our own browser automation inhouse. um we uh lean on kernel for the browser itself, but we do all of the execution and the logic and the handling uh in-house itself. [clears throat] Uh we've layered in a few important pieces for that like persistent sessions. So if we're logged in, we can um uh you know just if you if a user logs in once, they can actually expedite the checkout process the second time. Um and uh we allow you to save items to your cart, come back later. There's all kinds of like really nice configurations that you can build as a developer with what we have.

Now you can see our agent is going through the process here on the browser filling in all the information and you know on average we're seeing this take about 60 seconds or so. So it's coming back pretty quickly. Now it has a quick question for me. It also comes back and will ask you um you know if it gets stuck it doesn't know how to answer something. It will return back to you and and have you fill it in. Once again, I could save my card on file or tell it my phone number and then I wouldn't even have to be human in the loop on this. [clears throat] The final question is like where is this all going?

As I mentioned earlier, very very very few merchants have released anything in the way of UCP or ACP integrations. The [clears throat] really big like hurdle that's going to happen over the next few years is that merchants are going to require things like know your agent identification and anti-fraud um trusted agent protocols. So when your agent shows up to a merchant in the future, it's going to need to identify itself and say, "Hey, I'm Alex's agent. Here's why you can trust me, and here's my, you know, W3C credential and some of the progressive disclosures that I'm handing out." uh in order for you to believe all this [clears throat] um we are integrating all of that into the crossman checkouts API so that as a developer you don't have to think about

it the idea is that um while we're handling this through browser automation today and you can see we've you know handled this order here in about a minute um in the future we are just going to orchestrate across the protocols so the agent's going to show up and buy something in order to buy something it doesn't need browser automation it needs to say find me the UCP endpoint point. Let me identify myself as a as Alex's agent and whatever other signals the merchant needs and here we go. Let me buy. All right. I've been talking for a long time. We purchased something on the way. What questions do you have?

>> Absolutely. Yeah. Uh just just wanted uh to confirm that this was like a legitimate purchase online. You're going to you're going to get these gloves, right? >> Oh, yeah. I'm gonna get these gloves. I'll show you the receipt here. Um, probably came to my uh Yeah. Yeah. Let's Let's show you the order confirmation. Got a got a flip flip uh windows here. There they are. And as you can see, it just came in. Um, you know, zero minutes to go here.

>> Look at that. Yeah, that was like a legitimate like end to end um end to end flow folks for folks watching and and not kind of realizing what what just happened. Uh an agent >> went and bought uh a pair of gloves uh for Alex on his behalf. U made a real payment. Um hello hello, welcome to the future. [laughter] Um and and yes um like uh like you mentioned there was kind of some parts in there that were that were like human in the loop where you had to accept >> accept the the actual payment verify verify some things there. Um but yeah, eventually this this is kind of go >> Yeah.

>> Well, in the autonomous fashion where like maybe you just have like like you mentioned, you have like a set uh set thing a set of things you want to buy like the multivitamins or or like toothpaste or whatever, right? It's just like you set it and forget it. And yeah, that's >> where you pre-authorize a certain budget for the agent. you say, "Hey, look, you know, you're this is my shop, my clothing budget, and you have 300 bucks a month. If I tell you to buy something, just buy it." You know, I don't need to approve every single purchase individually. Um, we're definitely headed in that direction. And I think, you know, one of the other interesting ideas for for everyone out there who's building like a um an agent that reads your email, I mean, like what better way to understand a consumer the consumer

behavior than like, hey, just go parse like three years worth of emails and see what this person orders and like you could have an agent that very very quickly makes super intelligent recommendations. Like you could have it say, "Hey, okay, based on your spending, I need this budget. I'll [clears throat] take care of your, you know, your day-to-day like groceries and everyday household needs. And, you know, maybe I'll make some recommendations for the clothing that you like or the hobbies that you have, like whatever it is. >> Absolutely. And that that's a that's a startup idea. So, if anyone anyone was stuck on on what to build next, boom, there you go. Uh, free free ideas.

[laughter] Uh and um you kind of you kind of mentioned it as well um in terms of the um like UCPs, ATAS, AP2s. Um like what in your mind is kind of like the blocker or inhibitor of u merchants like kind of activating activating that layer? Um I I know I know like a lot of this stuff is is still being developed and evolved but um >> can you point it to like any particular thing like why some people are slow to move?

>> I think there'll be a lot of people who would say oh that you know this protocol is not fully threat uh fleshed out and there are some things that aren't thought through and you know we could we could point out technical holes. I think it's purely behavioral and I think that there are you know especially larger merchants like there's just you know they can wait they can afford to wait and and one of the stats I point to is um only 2.5% or something like that of US households have a subscription uh to an an agent or to an LLM and you know we are still very early in the adoption curve of these tools just generally. So even though you

and I are on the technical cutting edge and you know love to live and breathe this stuff [clears throat] before the critical mass is like really begging for these >> absolutely absolutely it's just a matter of time and um yeah the these these merchants need to realize that like AI agents these agents are going to be an emerging customer base of theirs. um maybe not today but um definitely tomorrow. So uh being able to kind of prepare for that and kind of build for that I think it's vitally important uh regardless of of of the size. Um so definitely um if this uh intersection of

payments and AI is something that interests you um all those resources um for you to get started will be in the description of the podcast below. Um, and with that, uh, Alex, just want to thank you so much for taking the time out of your busy day, uh, to come chat with us. Uh, I just want to wish everybody a very happy Friday, happy weekend, wherever you may be, and we will catch you here, uh, next week for another great episode of Devantel. Uh, till then, have a good one, folks. Cheers. If you enjoyed this episode of Devantel, be sure to subscribe to never [music] miss an episode. Till the next one cheers.