Agent Spending Without Controls — Rodrigo Coelho & Pranav Maheshwari, Edge & Node

Read the talk

Agent Spending Without Controls: Paid Tools, Checkout and Wallet Screening

Selected presentation frame from Agent Spending Without Controls — Rodrigo Coelho & Pranav Maheshwari, Edge & Node at 1120 seconds
Agent Spending Without Controls: Paid Tools, Checkout and Wallet Screening

Rodrigo Coelho and Pranav Maheshwari explain Ampersend through three demonstrations: buying access to information, delegating a gift purchase and rejecting a flagged wallet. The central question is how to give agents purchasing power while enforcing the rules that make that power acceptable.

From a talk by Rodrigo Coelho and Pranav Maheshwari

At a glance

Ideas worth remembering

  • Agent payment infrastructure must address both transaction economics and authorization. Coelho describes batching to reduce gas fees for tiny payments, while the final demonstration shows a screening rule determining whether a wallet may transact.

  • The contact lookup illustrates how paid-tool access can change task completion: Maheshwari reports an email format without the skill file and specific contact details with it, backed by an authorized lookup transaction. It does not establish the returned details’ accuracy.

  • Checkout completion does not establish budget enforcement. The gift request specifies less than $10, but the reported $9 total, $11 charge and $10 gift leave the final amount and compliance with that limit unresolved.

  • The simulated screening test demonstrates a concrete policy outcome: both wallets authorize with screening off, and the flagged wallet is rejected as blocklisted after screening is enabled. This supports that specific enforcement behavior, rather than a guarantee of comprehensive compliance.

  • The broader design combines tool access, remembered user context, payment capability and policy enforcement. The speakers argue that enterprises need these controls because a human officer remains accountable for autonomous spending, even when each transaction happens without human review.

From query micropayments to agent commerce

Selected presentation frame from Agent Spending Without Controls — Rodrigo Coelho & Pranav Maheshwari, Edge & Node at 176 seconds
From query micropayments to agent commerce

Rodrigo Coelho introduces Ampersend as a project incubated within Edge & Node to address the missing infrastructure around agent payments. He places that work in the context of The Graph, the blockchain data indexing protocol his team built. According to Coelho, the protocol has operated since 2018, and its decentralized network has served 1.8 trillion queries of onchain data to applications. That background matters because paying for individual data requests was already a practical concern before agents became the focus.

Coelho describes a progression from credit card payments to online commerce and then to agents that can transact. The engineering change is that a software system can now choose to purchase something as part of completing a task. Edge & Node’s earlier work provides a concrete example: Coelho says the team developed a query micropayment system in 2021 and referenced the HTTP 402 specification at that time. In late 2024, it began examining how agents could interact with The Graph’s data and pay for queries.

He then describes collaborating with Coinbase and Google, joining the foundation around x402 and contributing earlier micropayment and batching work to the specification. The stated purpose of batching is to reduce gas fees for very small payments. This exposes an economic constraint: a payment mechanism must keep its overhead low enough to make tiny purchases useful. Coelho also mentions Circle’s nano payments work and a growing ecosystem of payment protocols, infrastructure and governance. He does not explain the batching algorithm or quantify its savings, so the supported technical point is its intended role in reducing transaction costs.

0:150:20
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

0:01 · section reference included

Machine-speed payments still need accountable decisions

Coelho’s next concern is the decision process surrounding a payment. Traditional financial systems, he argues, assume a human somewhere in the loop deciding whether a transaction may proceed. Agents can transact at machine speed and around the clock, so controls designed around human activity need a different method of enforcement. He describes increasing experimentation with retail payments but says enterprise adoption remains early. Transaction growth and technical experimentation therefore do not, in his account, establish that organizations are ready to authorize autonomous spending.

The missing layer must answer questions about counterparties: whether an entity is sanctioned, whether it has been involved in terrorist activity and who is behind it. A wallet address supplies a destination for a transaction but, as Coelho emphasizes, may arrive without background information or identity. His proposed compliance layer would supply the checks needed before enterprises feel comfortable adopting the technology. The presentation identifies these requirements without specifying how identity is established or how every category of risk would be assessed.

Responsibility ultimately remains with a person, such as a chief legal officer or chief policy officer, who must sign off on deployment. Coelho says that person needs confidence that agents will not hallucinate, overspend or break policy. He cites potential fines reaching billions of dollars to explain why financial services organizations care about these controls. This is a statement of the assurance enterprises want, rather than evidence that the proposed system eliminates every failure. He calls Ampersend’s approach agentic checkout with a financial harness and hands the presentation to Pranav Maheshwari for the demonstrations.

3:333:36
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

3:33 · section reference included

The same lookup produces different results

The first task asks both environments to find email information for the head of crypto and blockchain at Mastercard. Maheshwari says the environment without the skill file cannot provide the exact address and instead returns the company’s email format, leaving the remaining work to the user. He reports that the environment with paid tools supplies the specific email address, a Twitter handle and the person’s location. The example connects task completion to access to an information service: the payment-enabled environment can buy a lookup that the other environment does not have available.

Maheshwari explains that the aggregator handles payment in the background. He points to a transaction made to obtain the requested information from a paid endpoint, then moves on while navigating the transaction view. The mechanism he describes is a task leading to a paid service call, with a corresponding spending record. The supplied account does not establish the contact information’s accuracy or give an unambiguous price for this lookup. Its supported result is the reported difference in specificity and the associated payment activity.

9:439:45
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

9:43 · section reference included

Delegating a gift purchase exposes the budget question

The second demonstration moves from purchasing information to purchasing a physical gift. Maheshwari asks the agent to use Shopify UCP to buy his father a Father’s Day gift costing less than $10. He describes the intended sequence as finding shops, identifying suitable gifts, listing options and buying through the Ampersend wallet. This combines product discovery with checkout: the agent needs both commerce tools and a payment route to carry the request through to an order.

While the shopping task runs, he returns to the earlier lookup and identifies its transaction as authorized. He uses it to reinforce the benefit of staying in the terminal rather than separately arranging access to paid services. Returning to the gift, he says the agent already retains information such as his name, address, phone number and preferences, so he does not need to enter those details repeatedly. The proposed convenience comes from combining remembered user context, tool access and payment capability in one workflow.

Maheshwari reports order confirmation, completion and a receipt, but the amounts remain unresolved. His account mentions a $9 total, an $11 charge and a $10 gift. These figures do not establish that the completed purchase respected the request to stay below $10, and no explanation reconciles them. The demonstration therefore supports a reported checkout completion while leaving budget enforcement uncertain. A spending instruction and a successfully completed order are separate claims; this example does not prove both.

11:3211:40
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

11:32 · section reference included

A simulated counterparty test separates payment from permission

Selected presentation frame from Agent Spending Without Controls — Rodrigo Coelho & Pranav Maheshwari, Edge & Node at 1032 seconds
A simulated counterparty test separates payment from permission

The final demonstration returns to compliance. Maheshwari argues that merchants need confidence about the origin of payments, and that both buyer and seller sides need a compliance layer. He sets up two agents: one uses an ordinary wallet, while the other uses a wallet flagged for the simulation. He describes the second as sanctioned but also explicitly allows that its flagged status was simulated. The experiment consequently tests behavior against a configured risk signal; it does not establish the wallet’s actual sanctions history.

Maheshwari first disables screening and saves the setting. He expects both agents to be able to transact in that state, including the one using the flagged address. This establishes the control condition: the payment system can accept a transaction without applying the wallet-screening rule. Whether a payment can be processed and whether policy permits it are distinct decisions in the demonstration.

16:1116:14
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

16:11 · section reference included

Screening rejects the flagged wallet

Selected presentation frame from Agent Spending Without Controls — Rodrigo Coelho & Pranav Maheshwari, Edge & Node at 1115 seconds
Screening rejects the flagged wallet

The seller in the test offers a website-scraping service that charges 0.1 cent for access to the scraping information. With screening disabled, Maheshwari reports that transactions from both wallets are authorized. The small price makes the example relevant to paid tool calls: a compliance decision may be needed even when the individual purchase is tiny. Authorization in this first state shows that the payment path works, while the flagged address is still allowed through.

He then enables the wallet-scanning feature built with TRM. Maheshwari explains that the ordinary wallet should continue working while transactions from the flagged wallet stop. He reports rejected and denied transactions and identifies the reason as the wallet address being blocklisted. The demonstrated mechanism is a screening result changing the authorization outcome. That is narrower than comprehensive compliance: the example shows enforcement of a wallet blocklist decision, without establishing identity resolution, coverage of every policy or protection against every form of agent misconduct.

Maheshwari closes by predicting that agents will use payment arrangements built around wallets or their own credit cards, rather than relying on familiar human checkout flows. He again connects agent usefulness to paid MCP services and the infrastructure needed to purchase them, and ends by inviting discussion about Ampersend. The ending preserves the presentation’s practical direction: make paid capabilities accessible from the agent’s workflow, while giving the payment system a way to refuse transactions that violate an enforced rule. The predicted growth of paid tools remains a forecast.

17:4217:43
Suggest correction

This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.

17:42 · section reference included

Read the complete timestamped transcript
  1. 0:01

    [music]

  2. 0:12

    All right, thanks for having me. Uh,

  3. 0:14

    today we're going to be speaking about

  4. 0:15

    ampersend and agent spending without

  5. 0:18

    controls, which is the missing

  6. 0:20

    infrastructure layer for AI payments. A

  7. 0:24

    little bit about myself, uh, I'm CEO of

  8. 0:26

    Edge and Node. For those that aren't

  9. 0:28

    aware, we were the team that built the

  10. 0:30

    graph protocol. For those that aren't

  11. 0:32

    aware of that, um it is a blockchain

  12. 0:35

    data indexing protocol. It's been around

  13. 0:37

    since 2018. We have a decentralized

  14. 0:39

    network. Um we've served 1.8 trillion

  15. 0:43

    queries over the years of onchain data

  16. 0:45

    to applications um utilizing blockchain

  17. 0:50

    data. Um I've been a serial entrepreneur

  18. 0:54

    myself and we've incubated a project

  19. 0:56

    within engine node called ampersend uh

  20. 0:59

    in the ejected commerce space which is

  21. 1:01

    what we're going to be talking about

  22. 1:02

    today.

  23. 1:04

    And so we're all aware that payments

  24. 1:06

    have been online for the past 30 years.

  25. 1:09

    Um you know starting in 1981 we had our

  26. 1:12

    first electronic transaction

  27. 1:14

    and fast forward we had a multi-deade

  28. 1:16

    transition. Um, and this is from credit

  29. 1:19

    card payments to online e-commerce

  30. 1:22

    payments. Uh, fast forwarding all the

  31. 1:24

    way to today where we have the rise of

  32. 1:27

    the agent economy and unlocking a new

  33. 1:29

    wave of economic value.

  34. 1:32

    And now we have LLMs and agents

  35. 1:36

    uh that can transact

  36. 1:38

    and we have this underlying payment

  37. 1:40

    infrastructure that's been around for

  38. 1:42

    decades and it's being rebuilt today um

  39. 1:45

    for the agentic economy. we've heard

  40. 1:47

    from many speakers here about how that's

  41. 1:49

    being done. We're going to be discussing

  42. 1:51

    today um our perspective and angle on

  43. 1:54

    it. So we we ourselves at edge and node

  44. 1:58

    were really early into the agentic

  45. 2:00

    commerce world. We actually developed um

  46. 2:03

    a microp payment system for queries back

  47. 2:05

    in 2021. We already referenced the 402

  48. 2:08

    spec back back then in a blog post. And

  49. 2:11

    so back in late 2024, we were looking

  50. 2:13

    into agents and how they would interact

  51. 2:17

    with data, how they would pay for it,

  52. 2:18

    specifically for the graph protocol

  53. 2:20

    itself and making microp payments for

  54. 2:22

    queries. And in our research, we were

  55. 2:25

    kind of looking at how this this could

  56. 2:28

    happen. And then a few weeks later, X42

  57. 2:30

    was released by Coinbase. So we jumped

  58. 2:33

    on immediately with the Coinbase team,

  59. 2:35

    with the Google team, started to

  60. 2:37

    collaborate to the X42 spec. We've

  61. 2:39

    joined the foundation. We've contributed

  62. 2:41

    to the X42 uh spec ourselves leveraging

  63. 2:44

    some of our prior art on the um microp

  64. 2:47

    payment system and uh batching protocol

  65. 2:50

    as part of how it could be utilized to

  66. 2:53

    uh reduce gas fees when you're dealing

  67. 2:55

    in nano payments. Um Circle themselves

  68. 2:58

    came out with their own kind of version

  69. 2:59

    of it called nano payments. Um so we

  70. 3:02

    were also working in similar veins

  71. 3:05

    and so we've seen an explosion of uh

  72. 3:09

    companies emerge solutions technologies

  73. 3:11

    and this kind of agented commerce map

  74. 3:14

    you can see on the screen here that

  75. 3:16

    there's a vast array of companies more

  76. 3:18

    added every day. It's really burgeoning.

  77. 3:20

    Um, and you know, we're just one little

  78. 3:24

    piece in here of people developing, you

  79. 3:26

    know, infrastructure,

  80. 3:28

    um, payment protocols, governance, etc.

  81. 3:33

    And so, traditional payment rails though

  82. 3:36

    were built for humans. Um, you know,

  83. 3:38

    when we're dealing with like financial

  84. 3:39

    institutions specifically, there's kind

  85. 3:41

    of a human in the loop that sets uh the

  86. 3:44

    decision making process of uh whether

  87. 3:46

    that payment is allowed to go through.

  88. 3:48

    But we're dealing with agents that

  89. 3:50

    transact at machine speed and around the

  90. 3:52

    clock. And like I was saying is all

  91. 3:54

    these controls and uh policies and rules

  92. 3:57

    were built for humans and not for uh uh

  93. 4:02

    machines that don't breathe, right? So

  94. 4:04

    we're dealing in um in microscond

  95. 4:07

    machine speed and it's just simply not

  96. 4:10

    going to work in the way it has. And so

  97. 4:11

    we need a new method of doing that. Um

  98. 4:14

    you can see a lot of the transactions uh

  99. 4:16

    exploding. This is sort of over time.

  100. 4:18

    We're seeing um transaction growth over

  101. 4:21

    X42. We're all saying this is coming. It

  102. 4:24

    is still early days. We're uh ourselves

  103. 4:27

    we're seeing a lot of experimentation

  104. 4:28

    kind of with uh with OpenClaw people

  105. 4:31

    making retail payments on an

  106. 4:33

    experimental basis. still not on the

  107. 4:36

    enterprise side as we're talking to uh

  108. 4:38

    many teams. Definitely everyone's

  109. 4:41

    looking at it and building kind of the

  110. 4:43

    infrastructure for when um these use

  111. 4:47

    cases break through. And so that's part

  112. 4:49

    of what we're we'll be displaying today

  113. 4:51

    with Amperson.

  114. 4:53

    But in order for this to really break

  115. 4:54

    through, we need a compliance layer

  116. 4:56

    specifically when you're dealing with

  117. 4:57

    like you know uh almost a quadrillion

  118. 5:00

    dollar industry of the the traditional

  119. 5:03

    financial world. there's a set of rules,

  120. 5:06

    policies, and guidelines that need to be

  121. 5:07

    in place before anything can happen. Uh

  122. 5:10

    beyond just a credit card transaction,

  123. 5:11

    we're dealing with trillions and

  124. 5:13

    trillions of dollars um being transacted

  125. 5:16

    um across the globe. And so we're going

  126. 5:19

    to be honing in a bit on that as part of

  127. 5:21

    the presentation today in the demo

  128. 5:23

    that's coming up. But the compliance

  129. 5:25

    layer is literally like is is this

  130. 5:29

    counterparty um a non-sanction entity?

  131. 5:32

    is this counterparty uh have they been

  132. 5:34

    involved in terrorist activity? Who is

  133. 5:36

    the identity behind this? And when

  134. 5:39

    you're dealing with a Gentic systems,

  135. 5:40

    it's just simply you're presented with

  136. 5:42

    just a wallet address with no kind of

  137. 5:45

    background information on that. And we

  138. 5:47

    need these uh infrastructure layers in

  139. 5:50

    place to um to enable and facilitate

  140. 5:53

    enterprises to feel comfortable to adopt

  141. 5:55

    this um technology. All of these systems

  142. 5:58

    exist in the existing world today. But

  143. 6:01

    uh as we're breaking through into this

  144. 6:03

    um agentic world, we really need to have

  145. 6:06

    all of these pieces uh before uh large

  146. 6:10

    enterprises are going to sign off and uh

  147. 6:14

    give the okay to um implement these. At

  148. 6:17

    the end of the day, there is going to be

  149. 6:19

    a human responsible like a chief legal

  150. 6:22

    officer, a chief policy officer and that

  151. 6:24

    will have to sign off and that person

  152. 6:27

    needs to feel 100% confident that the

  153. 6:30

    systems in place will not allow for um

  154. 6:34

    agents to hallucinate, to go off the

  155. 6:37

    rails, to overspend, um to break policy.

  156. 6:41

    Uh a lot of these compliance issues deal

  157. 6:44

    with large fines into the tens,

  158. 6:46

    hundreds, even billions of dollars. So

  159. 6:49

    these are really important um items and

  160. 6:53

    things to think about from a large

  161. 6:55

    enterprise and financial services

  162. 6:57

    perspective

  163. 6:59

    and governance simply hasn't caught up.

  164. 7:01

    We're again we're early days so we are

  165. 7:04

    part of building up what needs to be put

  166. 7:06

    in place from our view in terms of

  167. 7:09

    allowing this to scale.

  168. 7:11

    And that's why we're building Amperent.

  169. 7:14

    Um, so we're dealing with Agentic

  170. 7:17

    checkouts with a financial harness.

  171. 7:20

    And I'm going to turn it over to my

  172. 7:22

    colleague Prrenoff who is going to show

  173. 7:24

    us a demo. Right.

  174. 7:25

    >> All right. Thanks.

  175. 7:32

    >> All right. Like we said, for agents to

  176. 7:35

    be really useful, we need to give them

  177. 7:38

    tools. Right now they're being used for

  178. 7:40

    coding. If you want to use them to build

  179. 7:43

    better things, you need to give them the

  180. 7:44

    right tools. As of right now, as a AI

  181. 7:47

    industry, we've given them a lot of MCP

  182. 7:49

    servers,

  183. 7:51

    but most of these MP MCP servers are

  184. 7:54

    free of cost. So there are two ways to

  185. 7:57

    make your agents really useful.

  186. 8:00

    One way is you can go to all these

  187. 8:02

    websites like Exop, Firecrawl and so

  188. 8:06

    many more which are getting more and

  189. 8:08

    more bigger with more MCP servers coming

  190. 8:11

    to play and you got to put in your

  191. 8:12

    credit card and you make your agent

  192. 8:14

    better or there is a better way that you

  193. 8:18

    can use an aggregator which has all the

  194. 8:21

    important tools to make your agent super

  195. 8:24

    powerful and that's why we built this

  196. 8:28

    marketplace and all you need is a skill

  197. 8:31

    file installed in your agent. You can

  198. 8:33

    give your cloud code this skill file and

  199. 8:36

    everything else will be done for you. So

  200. 8:39

    I'm going to show you an experiment.

  201. 8:41

    We'll have uh this specific terminal

  202. 8:44

    with the amperson skill file and we'll

  203. 8:47

    have cloud code cloud co-working without

  204. 8:51

    the skill file and you'll see the

  205. 8:54

    difference is immense already. The

  206. 8:56

    reason is that most important MCP

  207. 8:59

    servers are going to be paid and you'll

  208. 9:01

    not put in your credit card to all these

  209. 9:03

    MCP servers. Rather what you'll do is

  210. 9:06

    use you'll use the agentic commerce tool

  211. 9:08

    or a platform or a wallet or a credit

  212. 9:10

    card to make that happen. And that's

  213. 9:12

    what what we enable. So let's just go on

  214. 9:16

    the terminal and install the ampus and

  215. 9:19

    scale file in one set. This is our

  216. 9:21

    website by the way where you can go. You

  217. 9:24

    copy the prompt and just like put in

  218. 9:26

    here. I already have it so I'm not going

  219. 9:28

    to uh sort of maybe it's fine. Uh

  220. 9:32

    install the skill file. And here I'm on

  221. 9:35

    the cloud coork, right? It does not have

  222. 9:38

    the skill file. Fair and square. This

  223. 9:41

    has the skill file. Now what I'll do is

  224. 9:43

    I'll tell them find the email

  225. 9:45

    information of the head of crypto and

  226. 9:47

    blockchain at Mastercard. All right. And

  227. 9:49

    I'll put in the same prompt. Uh I'll put

  228. 9:53

    in the same prompt for cloud cowork.

  229. 9:56

    Remember the only difference is this

  230. 9:59

    specific terminal of mine has the skill

  231. 10:03

    file which has paid MCP tools and this

  232. 10:06

    one does not. And you'll see that it is

  233. 10:09

    not able to like specifically find out,

  234. 10:11

    hey, we're not able to give you exactly

  235. 10:13

    the email, but this is how the email

  236. 10:15

    format is for our master card. And you

  237. 10:18

    can go ahead and find that out. But when

  238. 10:20

    I go to my terminal, it is already be

  239. 10:22

    able to it will already be able to give

  240. 10:24

    me the email specifically the Twitter

  241. 10:27

    and what where is he based and

  242. 10:29

    everything and beyond. This is just one

  243. 10:32

    example of telling you that agents are

  244. 10:34

    getting powerful through MCP. But these

  245. 10:37

    MCPs are not public services. They're

  246. 10:40

    going to be charged and you will not

  247. 10:42

    know that you're interacting with MCPS

  248. 10:43

    or installing it or going on their

  249. 10:45

    websites and getting the credit card.

  250. 10:47

    Rather, you'll just have a scale file

  251. 10:49

    which is a aggregator which can take

  252. 10:51

    care of your payments. in the

  253. 10:54

    background. What happened was if you go

  254. 10:56

    over here, you already can see that we

  255. 10:59

    did a transaction so that we could

  256. 11:02

    enable this surge for you because this

  257. 11:04

    is a paid endpoint that we had to

  258. 11:06

    specifically pay for to get what you

  259. 11:09

    specifically needed. So when we go to

  260. 11:11

    transactions, you should be able to find

  261. 11:13

    that we did a small transaction so that

  262. 11:16

    you would specifically get that specific

  263. 11:19

    information.

  264. 11:21

    uh 30 days on the

  265. 11:24

    maybe we skip to the next demo for now

  266. 11:27

    as you can see this spending happened.

  267. 11:30

    Let's skip to the next demo. I wanted to

  268. 11:32

    show you one more demo. In this what

  269. 11:34

    I've done is let's say I'm not a good

  270. 11:38

    kid which generally I am and I want to

  271. 11:40

    buy a father's day gift for my dad and I

  272. 11:44

    can also make that happen directly via

  273. 11:46

    the terminal. Shopify introduced UCP.

  274. 11:50

    Amazon might be coming up with its own

  275. 11:52

    thing, but what I'll do is I'll give it

  276. 11:54

    a specific command using Shopify UCP.

  277. 11:57

    Buy my father a Father's Day gift. Keep

  278. 12:00

    the gift less than $10. Let's see how

  279. 12:03

    this specific thing rolls.

  280. 12:06

    What uh will happen in this specific

  281. 12:08

    terminal case is that it will already be

  282. 12:11

    able to f locate certain shops online

  283. 12:16

    and find out what things can be given

  284. 12:19

    for father's day gift list them for me

  285. 12:22

    and be able to buy that directly via

  286. 12:25

    ampend wallet that we have created. your

  287. 12:28

    agent is as powerful as the paid MCP

  288. 12:31

    tools that you're connected to it and if

  289. 12:34

    you've given it a payment trail.

  290. 12:38

    So currently it's finding that out for

  291. 12:40

    us.

  292. 12:43

    Okay.

  293. 12:54

    See, this was an authorized transaction

  294. 12:56

    it did for finding me the email of that

  295. 12:59

    specific person at Mastercard, right?

  296. 13:02

    And these are all paid services or I

  297. 13:04

    would have to go on Xi and sort of put

  298. 13:07

    in my email card and get that. So, never

  299. 13:10

    leave your terminal. Just have a payment

  300. 13:12

    gateway and paid MCP tools and make your

  301. 13:16

    agents super super compatible and faster

  302. 13:19

    and give them more uh power.

  303. 13:24

    All right, it's currently finding some

  304. 13:27

    stuff for me. Let's see if it's able to

  305. 13:29

    come up with some recommendations that I

  306. 13:31

    can use and I can hear itself do a

  307. 13:35

    aentic checkout so I can buy something

  308. 13:37

    for my dad for Father's Day.

  309. 13:42

    What's good about this specific thing is

  310. 13:45

    that again and again I don't have to put

  311. 13:46

    in my name, my address, my phone number,

  312. 13:50

    anything. The agent already has the

  313. 13:53

    memory where it knows what do I like,

  314. 13:56

    what do I not like, and it will just

  315. 13:58

    give me specific things that I might

  316. 14:00

    want to buy for him. So, it already

  317. 14:04

    like, you know, total $9. This is what

  318. 14:07

    it is. Confirming the order and it

  319. 14:09

    already placed the bet and it already

  320. 14:11

    took $11 for that

  321. 14:16

    order complete. And here is your

  322. 14:18

    receipt. and my dad gets his $10 key

  323. 14:23

    directly via the agent. So this is just

  324. 14:27

    starting of what we call will be the

  325. 14:29

    future of agent e-iccommerce. It's not

  326. 14:32

    just buying gifts for your uh parents

  327. 14:34

    but rather it's much bigger. MCPs are

  328. 14:38

    becoming the normal norms of how your

  329. 14:40

    agents become super powerful and you

  330. 14:44

    might have already seen that Cloudflare

  331. 14:47

    is opening its gateway through X42 and

  332. 14:50

    agentic payments. If you go on a website

  333. 14:53

    and the agent is crawling that website

  334. 14:56

    then ads are irrelevant. How do people

  335. 14:58

    get paid?

  336. 15:00

    It's via agentic payments where a bot

  337. 15:03

    comes through Cloudflare, pays a

  338. 15:05

    microtransaction and gets all the

  339. 15:08

    information that's needed. So to make

  340. 15:11

    your agent super powerful, you need paid

  341. 15:14

    tools. We are in the era where most MCP

  342. 15:17

    tools are free. That's not going to be

  343. 15:19

    the case in the future. they're going to

  344. 15:21

    get paid and it will be it what your

  345. 15:24

    agent would need is good MCPS or an API

  346. 15:28

    which has an integration of MCPS and a

  347. 15:31

    wallet and this is what I'm trying to

  348. 15:34

    show you.

  349. 15:35

    So the specific bet has already been

  350. 15:38

    placed. I'm going to try to show you the

  351. 15:40

    transaction over here that was done and

  352. 15:44

    it's already settled and I might have

  353. 15:46

    gotten an email where Shopify was able

  354. 15:50

    to buy my dad a gift. Let's see if I

  355. 15:54

    open it and show it to you because it

  356. 15:56

    Yeah, it's okay. You guys are friends.

  357. 16:00

    See over here the order has been placed

  358. 16:02

    and I can already it's already getting

  359. 16:05

    tracked all via the agent terminal. I

  360. 16:08

    didn't have to even skip. Let's go to

  361. 16:11

    the last part which is all of this will

  362. 16:14

    be big if and only if like Rodrigo says

  363. 16:18

    there is compliance. Merchants will not

  364. 16:20

    take payments if they think this order

  365. 16:23

    is being placed by North Korean wallet.

  366. 16:26

    And for that you need the compliance

  367. 16:28

    layer to be fitting in as well. The

  368. 16:30

    seller and the buyer side both need to

  369. 16:32

    have that. And I've created this

  370. 16:34

    specific uh simulation for you in which

  371. 16:39

    we have a good claw that you can let me

  372. 16:41

    just close this right now. So now let's

  373. 16:44

    go to the third one which is compliant

  374. 16:47

    transactions

  375. 16:49

    right for the world and Amazon to accept

  376. 16:51

    agentic payments there needs to be

  377. 16:53

    compliance involved to make sure that

  378. 16:56

    the payment that's been done is not

  379. 16:57

    malicious. So we have a good claw that

  380. 17:00

    I'll spinning up that I'll be spinning

  381. 17:02

    up over here and we have a bad claw. The

  382. 17:05

    thing about bad claw is this is again a

  383. 17:07

    simulation but this is a sanctioned

  384. 17:10

    wallet address that means it has either

  385. 17:13

    interacted with North Korean entities or

  386. 17:15

    we were able to simulate it in a way

  387. 17:17

    that it is flagged right so no amount of

  388. 17:20

    transaction should be able to be done

  389. 17:22

    because it is out of the swift scope of

  390. 17:24

    policies.

  391. 17:26

    So if we go over here uh in this

  392. 17:29

    specific thing and I disable screening

  393. 17:31

    and save it,

  394. 17:33

    my good claw and my B bad claw both

  395. 17:37

    should be able to do transactions.

  396. 17:42

    Just to give you a little bit of

  397. 17:43

    periphery, this is the seller that is

  398. 17:46

    accepting payments. We have a small

  399. 17:48

    service of scraping websites and it

  400. 17:51

    needs to pay 0.1 cent to be able to take

  401. 17:54

    that scraping information. Right now

  402. 17:57

    because we don't have regulations both

  403. 18:00

    these transactions are going through and

  404. 18:02

    you can see it's getting authorized over

  405. 18:04

    here. Right now what I'll do is using

  406. 18:08

    compliance I will enable the specific

  407. 18:11

    feature that we have built with TRM

  408. 18:13

    which scans the wallet and make sure

  409. 18:16

    that all the transactions are compliant

  410. 18:18

    and what it will do is that up till now

  411. 18:21

    everything has been working for good

  412. 18:23

    claw and bad claw but things will stop

  413. 18:26

    working for bad claw because using

  414. 18:29

    compliance we are able to scan that and

  415. 18:31

    we are able to give you see over here

  416. 18:33

    the your your transactions are getting

  417. 18:35

    blocked or rejected.

  418. 18:38

    And I can also show you over here that

  419. 18:40

    rejected and denied. And the reason

  420. 18:42

    behind the denying is because you're in

  421. 18:45

    the block listed wallet addresses.

  422. 18:48

    All to tell you that agentic commerce is

  423. 18:51

    becoming real. Agents need commerce more

  424. 18:55

    than humans need commerce. But it will

  425. 18:59

    not be the same. You will not go through

  426. 19:02

    payment guardrails of Stripe or any

  427. 19:06

    other of these checkout flows, but

  428. 19:08

    agents will have their own proprietary

  429. 19:10

    firms either through wallets or through

  430. 19:12

    their own credit cards. Agents would

  431. 19:15

    need to be superpowered. MCPs will get

  432. 19:18

    paid and more and more paid MCPs will

  433. 19:21

    come to reality to make your agent

  434. 19:23

    successful.

  435. 19:25

    That will only happen and only be

  436. 19:28

    successful in UX terms and more if

  437. 19:32

    you're able to empower your wallet with

  438. 19:34

    paid MCPs and a wallet infrastructure.

  439. 19:38

    If you want to know more about it, come

  440. 19:41

    to Ampend. We'll be here outside

  441. 19:44

    chatting if you have any questions.

  442. 19:46

    Thank you.

  443. 20:01

    >> [music]