WEBVTT

NOTE Sentence-level transcript of https://www.youtube.com/watch?v=G7cgLjZtmMU

NOTE One cue per sentence. Cue ids are the line anchors on /transcripts/G7cgLjZtmMU.html. A cue ends where the next begins, or 2 s after its last word.

s1
00:00:01.309 --> 00:00:03.309
[music]

s2
00:00:12.720 --> 00:00:14.240
My name is Anup Priyo.

s3
00:00:14.240 --> 00:00:16.400
I'm a senior engineering manager at Best Buy.

s4
00:00:16.400 --> 00:00:25.320
And me and my team are working together right now to figure out what does Agentic Commerce mean and how can we meet our customers where they're at.

s5
00:00:25.320 --> 00:00:29.560
And the newest place that they're at is at Agentic Services.

s6
00:00:29.560 --> 00:00:34.880
I'm excited to give my talk today and well, what's what's my credentials?

s7
00:00:34.880 --> 00:00:43.200
Where ever since I was a young boy, I dreamed of high throughput inference, harnessing my tools within a context window, kept in check with evals.

s8
00:00:43.200 --> 00:00:45.560
Yeah, that's absolutely correct.

s9
00:00:45.560 --> 00:00:48.920
In 2003, all those things definitely existed.

s10
00:00:48.920 --> 00:00:50.160
I kid.

s11
00:00:50.160 --> 00:00:53.840
Uh over the last 1 year, uh we have been learning a lot.

s12
00:00:53.840 --> 00:00:55.040
Shopping isn't new.

s13
00:00:55.040 --> 00:01:02.880
Shopping is probably one of the most fun things one can do and one of the most essential things that people need to do ever since uh the economy existed,

s14
00:01:02.880 --> 00:01:05.760
but I have been super excited by it.

s15
00:01:05.760 --> 00:01:14.040
So, I'm going to talk about what are the things some of the things that I've learned and hopefully uh share the notes.

s16
00:01:14.040 --> 00:01:16.320
So, what is Agentic Commerce?

s17
00:01:16.320 --> 00:01:24.720
I'm not going to go over the broad definition again, but basically it's the idea that AI assists will help you with your shopping journey.

s18
00:01:24.720 --> 00:01:28.040
Shopping has different facets to it.

s19
00:01:28.040 --> 00:01:41.480
For instance, there's a discovery, there's figuring out the aspects of do I actually truly need it, understanding and deciding, there's loyalty, there's pricing, there's fulfillment, post fulfillment.

s20
00:01:41.480 --> 00:01:42.600
It's a lot.

s21
00:01:42.600 --> 00:01:55.960
And believe it or not, right now about 45% of all agent sessions that happen within major providers like chat.gbt.com and Google Gemini are related to shopping.

s22
00:01:55.960 --> 00:02:00.560
Maybe it's a little biased uh that I don't use it as much, but I'm an engineer.

s23
00:02:00.560 --> 00:02:06.560
But, the humans out there are using AI and talking to them to help with their shopping journey.

s24
00:02:06.560 --> 00:02:10.240
So, it's also not a binary, you know.

s25
00:02:10.240 --> 00:02:13.800
Uh right now, we're at that state of human in the loop.

s26
00:02:13.800 --> 00:02:16.760
The ideal state would be autonomous shopping.

s27
00:02:16.760 --> 00:02:18.840
You tell what you're excited about.

s28
00:02:18.840 --> 00:02:22.280
Your agents goes around, talks to different merchants.

s29
00:02:22.280 --> 00:02:24.480
Uh I'm originally from Bangladesh.

s30
00:02:24.480 --> 00:02:27.360
We haggle a lot with the merchants, too.

s31
00:02:27.360 --> 00:02:29.880
Maybe does that, negotiate, does the payment.

s32
00:02:29.880 --> 00:02:32.680
But, right now, we're in the human in the loop.

s33
00:02:32.680 --> 00:02:42.640
And the talk today is going to talk about the mental model of how that human in the loop is working right now, while also provide you with the architecture if you choose to extend it,

s34
00:02:42.640 --> 00:02:46.440
or show you the vision of how autonomous shopping might work.

s35
00:02:46.440 --> 00:02:50.640
So, this isn't our first the first attempt.

s36
00:02:50.640 --> 00:02:56.120
Uh even a year ago, there were people trying to figure out how can we automate this.

s37
00:02:56.120 --> 00:03:00.440
Even now, you can go probably download the Cloud Chrome extension.

s38
00:03:00.440 --> 00:03:13.200
Maybe you've used Atlas, where you tell the AI you need something, you need headphones, you need uh that that grocery list of items that you have been meaning to buy, but never made the

s39
00:03:13.200 --> 00:03:16.959
actual effort to show up because, you know, you didn't have the time.

s40
00:03:16.959 --> 00:03:24.760
So, why don't you take screenshots, read the DOM, navigate to the merchant site, fill forms for me, do loyalty.

s41
00:03:24.760 --> 00:03:27.519
Kind of just didn't work as expected.

s42
00:03:27.519 --> 00:03:31.080
It was really clunky and slow and brittle.

s43
00:03:31.080 --> 00:03:45.040
And if you are a merchant who's trying to sell stuff, any engineering department of that merchant will tell you an AI impersonating or your browser is just firing up all the alarm bells.

s44
00:03:45.040 --> 00:03:56.960
So, a lot of times, you will probably be even stuck on the payment flow because we We want you to be using AI to put in that order, or at least in that phase, that's what was happening.

s45
00:03:56.960 --> 00:03:59.880
So, what did what did it actually work?

s46
00:03:59.880 --> 00:04:01.960
And is it actually working right now?

s47
00:04:01.960 --> 00:04:03.000
It is.

s48
00:04:03.000 --> 00:04:09.320
Chat GPT shopping, Google AI mode is doing just that.

s49
00:04:09.320 --> 00:04:20.040
Right now, agent tech shopping is considered to be a $7 billion industry and might go up to $65 billion industry by 2030.

s50
00:04:20.040 --> 00:04:30.240
And the majority of the shoppers are using the mainstream conversational AI assistants, which is on the browser or in your app, Chat GPT and Google Gemini.

s51
00:04:30.240 --> 00:04:33.200
We're also seeing that pop up in Instagram and Facebook.

s52
00:04:33.200 --> 00:04:36.320
Meta shop Meta wants to do meta commerce now.

s53
00:04:36.320 --> 00:04:41.240
I heard GoPuff and Grok came together to make an app as well.

s54
00:04:41.240 --> 00:04:49.280
And also Microsoft Copilot just yesterday announced in the UK that you can buy Ray-Bans now inside Microsoft Copilot.

s55
00:04:49.280 --> 00:04:59.640
So, to make that happen, Google and OpenAI separately came up with their own little primitives, ACP and UCP.

s56
00:04:59.640 --> 00:05:03.919
Which is basically talking about how you would actually talk to us.

s57
00:05:03.919 --> 00:05:12.480
For some of you who are shopping on the other side as the customer, there is not much of a difference between adding an item to cart, adding a second quantity.

s58
00:05:12.480 --> 00:05:15.600
But to us merchants, that's a second line item, buddy.

s59
00:05:15.600 --> 00:05:17.640
That's not the same scale.

s60
00:05:17.640 --> 00:05:25.960
So, if we don't talk about the nuances and the primitives of commerce and standardize it, things will just not work and will remain to be clunky.

s61
00:05:25.960 --> 00:05:33.880
So, ACP was Chat GPT's attempt at it, and Universal Commerce Protocol, UCP, was Google's attempt at it.

s62
00:05:33.880 --> 00:05:45.320
So, now that I've already established that this is happening, I just wanted to say that it is happening as easy as you go to the Chat GPT Gemini, tell it to find me cat cookies.

s63
00:05:45.320 --> 00:05:50.840
More to it why I chose cat cookies later uh in this example.

s64
00:05:50.840 --> 00:05:52.600
The AI surfaces the product.

s65
00:05:52.600 --> 00:05:55.000
Agent calls the merchant checkout API.

s66
00:05:55.000 --> 00:05:55.920
No browser.

s67
00:05:55.920 --> 00:06:01.200
Payment flows via scope payment mandate or a delegated payment token.

s68
00:06:01.200 --> 00:06:06.040
An order confirms and human kind of didn't have to touch the cart.

s69
00:06:06.040 --> 00:06:12.040
So, to all of this that's happening for the user, a lot is happening on the other side.

s70
00:06:12.040 --> 00:06:13.880
And it's kind of overwhelming.

s71
00:06:13.880 --> 00:06:19.919
One day we're talking about MCPs, another day A2A, ACP, UCP, AP2.

s72
00:06:19.919 --> 00:06:21.800
It's like what is even real?

s73
00:06:21.800 --> 00:06:25.880
Like if someone came up to me tomorrow and said, "I came up with HYPE."

s74
00:06:25.880 --> 00:06:28.919
I would probably think it's probably real.

s75
00:06:28.919 --> 00:06:32.840
So, I wanted to dissect this mental model for you as I've learned about it more.

s76
00:06:32.840 --> 00:06:37.919
MCP is still the model context protocol, the way that the AI agent identifies the tool.

s77
00:06:37.919 --> 00:06:42.640
So, maybe we can figure out what does this AI agent uh specifications are.

s78
00:06:42.640 --> 00:06:44.280
To showcase what products they have.

s79
00:06:44.280 --> 00:06:46.320
To showcase the details of a specific product.

s80
00:06:46.320 --> 00:06:48.120
To showcase loyalty.

s81
00:06:48.120 --> 00:06:50.360
A2A is how agents talk to each other.

s82
00:06:50.360 --> 00:06:52.160
They're more of a spec.

s83
00:06:52.160 --> 00:07:01.360
ACP, UCP are the primitives and AP2 is the agentic payment protocol scope payment mandate that Google's open specification came out.

s84
00:07:01.360 --> 00:07:03.960
And we'll talk all of them one by one.

s85
00:07:03.960 --> 00:07:06.560
How they actually relate to agentic shopping.

s86
00:07:06.560 --> 00:07:20.280
So, the MCP tool access is very important because without knowing the different capabilities and hitting those different capabilities, taking the time to bring it into context, understanding the user's memory,

s87
00:07:20.280 --> 00:07:24.040
the the agent will never be able to figure out what you're even trying to do.

s88
00:07:24.040 --> 00:07:28.600
And the only way to get access to the specific capabilities is through MCP tool calls.

s89
00:07:28.600 --> 00:07:31.000
Uh The next one is A2A.

s90
00:07:31.000 --> 00:07:34.680
So, now there are different ways to architect this.

s91
00:07:34.680 --> 00:07:38.080
Different capabilities I talked about like payments.

s92
00:07:38.080 --> 00:07:39.800
Uh let's say what do you call it?

s93
00:07:39.800 --> 00:07:40.480
Loyalty.

s94
00:07:40.480 --> 00:07:43.920
You can make agents about specific domains itself.

s95
00:07:43.920 --> 00:07:44.560
Sorry.

s96
00:07:44.560 --> 00:07:47.000
Uh Uh right?

s97
00:07:47.000 --> 00:07:51.480
And if you have specific domain level agents, agents need to talk to each other.

s98
00:07:51.480 --> 00:07:53.720
We need to find a standardized way to talk to each other.

s99
00:07:53.720 --> 00:07:58.600
So, A2A, the specifications kind of fill in that gap.

s100
00:07:58.600 --> 00:08:05.280
Uh also, if your customer agent and your merchant agent uh need to talk to each other, maybe you could They're both agents.

s101
00:08:05.280 --> 00:08:07.880
Maybe we can use A2A.

s102
00:08:07.880 --> 00:08:11.480
So, now to the UCPMCP primitives.

s103
00:08:11.480 --> 00:08:14.880
So, the most important data is that product data.

s104
00:08:14.880 --> 00:08:29.240
So, UCP allows for adding that product data in a more uh more organized way, and ACP does the same because we don't want to go through your PDP and crawl and figure out every specific attribute.

s105
00:08:29.240 --> 00:08:31.280
Merchant, just tell us.

s106
00:08:31.280 --> 00:08:33.560
And also, those pro- products change a lot.

s107
00:08:33.560 --> 00:08:37.240
So, maybe you can tell us when they change as well to send this.

s108
00:08:37.240 --> 00:08:43.360
So, that kind of data is happening uh uh uh that that kind of data flow is happening in the product feed.

s109
00:08:43.360 --> 00:08:46.839
Normally, you would assume that this would be a search catalog.

s110
00:08:46.839 --> 00:08:53.920
However, both ACP and UCP right now, so Gemini and ChatGPT does not support that search catalog call.

s111
00:08:53.920 --> 00:08:56.400
They want you to send that feed to them.

s112
00:08:56.400 --> 00:08:59.400
And for those of you who are like, "Why wouldn't you do that?"

s113
00:08:59.400 --> 00:09:00.480
There's reasons to it.

s114
00:09:00.480 --> 00:09:04.760
Uh sponsored products, retail media uh related things, ranking.

s115
00:09:04.760 --> 00:09:12.520
But, the most important technological challenges if you have M number of merchants and N number of products, now it has to call that many.

s116
00:09:12.520 --> 00:09:19.160
While if you send the product feed ahead of time, we can index that and be ready offload to offload when you ask for something.

s117
00:09:19.160 --> 00:09:22.080
The I've also put an example of Meta's uh product feed.

s118
00:09:22.080 --> 00:09:24.600
As you can see, they're similar, but still different.

s119
00:09:24.600 --> 00:09:25.880
Everyone has an opinion.

s120
00:09:25.880 --> 00:09:29.320
They think their opinion is the best one, and that's what they're rolling with.

s121
00:09:29.320 --> 00:09:32.880
So, there's three different specifications right here.

s122
00:09:32.880 --> 00:09:38.720
So, now that we talked about product feed, talking to to other, calling tools, let's talk about payments.

s123
00:09:38.720 --> 00:09:48.160
Uh right now, uh none of them are supporting that more autonomous form of, you know, X402 or some other kind of payments.

s124
00:09:48.160 --> 00:09:49.160
We're just not there yet.

s125
00:09:49.160 --> 00:09:50.560
We're just not confident yet.

s126
00:09:50.560 --> 00:10:02.280
We want more human in the loop, a merchant to be uh talking to a payment processor who will take the responsibility, or in this case, liability, to actually initiate the payments.

s127
00:10:02.280 --> 00:10:12.200
So, in chat GPT, payments only happen through a shared payment token right now, and Gemini UCP, the payments are only being accepted through Google Pay.

s128
00:10:12.200 --> 00:10:15.640
So, the scope demands will tell you what the products are.

s129
00:10:15.640 --> 00:10:22.080
What I'm excited about is more about AP2, which is an extension of UCP, which it You see what I'm talking about?

s130
00:10:22.080 --> 00:10:23.760
There's so many acronyms.

s131
00:10:23.760 --> 00:10:33.680
Uh AP2 is more about, "Hey, if we wanted to do autonomous, can you tell me who authorized the agent, what exactly can it buy, and what's the max amount

s132
00:10:33.680 --> 00:10:41.960
uh that should be able to haggle with, maybe, and then the revocation URL, and the user concept proof?"

s133
00:10:41.960 --> 00:10:42.480
All right.

s134
00:10:42.480 --> 00:10:43.360
Enough talking.

s135
00:10:43.360 --> 00:10:48.800
I love building stuff, so for the sake of this, I have put together a little demo.

s136
00:10:48.800 --> 00:10:55.800
For those of you who have remembered that cat cookie example, it's because the demo is about my cat.

s137
00:10:55.800 --> 00:11:03.320
Ginny is my orange tabby, and in this made-up example, Ginny has been has transformed into a bakery agent.

s138
00:11:03.320 --> 00:11:06.839
She wants to earn her keep by selling baked goods.

s139
00:11:06.839 --> 00:11:16.920
So, right now, the model that I'm using is from Cerebras at 3,000 tokens per second, so hopefully this will be really, really fast, and we can give you an example

s140
00:11:16.920 --> 00:11:18.440
of the entire flow.

s141
00:11:18.440 --> 00:11:24.160
And just like Chrome DevTools, I've kind of had a couple of tools in place to showcase what happens.

s142
00:11:24.160 --> 00:11:35.960
The first thing I will tell Ginny, my beautiful cat who's selling baked goods now, "Hi, tell me about uh all your products.

s143
00:11:36.880 --> 00:11:40.400
And this is supposed to be a demo, an example.

s144
00:11:40.400 --> 00:11:43.600
And uh Genie has given me exactly that.

s145
00:11:43.600 --> 00:11:46.080
All the different products that she might need.

s146
00:11:46.080 --> 00:11:47.360
So here, let's look at this.

s147
00:11:47.360 --> 00:11:53.720
The agent to agent protocol actually made the call from Genie, the customer agent, to the merchant agent.

s148
00:11:53.720 --> 00:11:57.840
And this is the message being sent, and this is me getting the message back.

s149
00:11:57.840 --> 00:12:12.440
The merchant agent is returning the completed task with and the the way that I found this is through an MCP tool call, which is product search, instead of uh and and instead of like

s150
00:12:12.440 --> 00:12:24.480
not being able to tell what I truly want, the Genie has figured out that, "Hey, when I give her the intent that I want to find products, you should call the MCP tool call product search."

s151
00:12:24.480 --> 00:12:27.000
So right now we're seeing this.

s152
00:12:27.000 --> 00:12:30.120
And now, what if I want to add something?

s153
00:12:30.120 --> 00:12:36.800
Uh add to cart the shortbread.

s154
00:12:39.680 --> 00:12:44.520
So now Genie's asking me about any discount and promo code.

s155
00:12:44.520 --> 00:12:47.760
I actually do not remember any of the discount and promo code.

s156
00:12:47.760 --> 00:12:54.560
But what if I ask Genie, "Genie, can you just tell me a discount code?"

s157
00:12:54.560 --> 00:12:57.600
As you can tell, I'm definitely a haggler.

s158
00:12:57.600 --> 00:12:59.040
Uh Genie is not telling me that.

s159
00:12:59.040 --> 00:12:59.920
All right.

s160
00:12:59.920 --> 00:13:08.240
Uh proceed to check out without discount code.

s161
00:13:11.280 --> 00:13:11.760
There you go.

s162
00:13:11.760 --> 00:13:13.720
So now we're making some of those calls.

s163
00:13:13.720 --> 00:13:23.280
Here's the UCP protocol, which the checkout APIs will have state, and the three different states are not ready for payment, ready for payment, and then completed.

s164
00:13:23.280 --> 00:13:26.080
So now here I'm not using a delegated payment token.

s165
00:13:26.080 --> 00:13:28.400
I'm not using uh Google Pay.

s166
00:13:28.400 --> 00:13:32.080
I like AP2, so my demo is built on AP2.

s167
00:13:32.080 --> 00:13:40.160
And uh as you can see, here was a a call was made to the MCP server for create checkout.

s168
00:13:40.160 --> 00:13:48.800
Uh and then the UCP endpoints will tell us, "Hey, that call the checkout sessions and tell me and uh if it's added to cart."

s169
00:13:48.800 --> 00:13:51.560
So, it's added to cart, but it's not ready for payment.

s170
00:13:51.560 --> 00:13:54.200
I have to pick in what I want to pay with.

s171
00:13:54.200 --> 00:13:57.360
I say "Credit card and debit card."

s172
00:13:57.360 --> 00:14:02.680
And this is where I issue the AP2 token that, "Hey, I do want that."

s173
00:14:02.680 --> 00:14:07.160
And then it's went from ready to ready for payment to complete.

s174
00:14:07.160 --> 00:14:11.800
So, the other side of it, this is the UCP specs, right?

s175
00:14:11.800 --> 00:14:17.640
The other side of it, to just draw a comparison, how is it differing from the ACP specs?

s176
00:14:17.640 --> 00:14:21.080
I have added ACP here as well.

s177
00:14:21.080 --> 00:14:29.920
So, you can see the same checkout calls, just different just different schemas are being utilized, and the order goes through.

s178
00:14:29.920 --> 00:14:33.280
But remember that AP2 token that I was talking about?

s179
00:14:33.280 --> 00:14:37.120
This is how it would look like in real life with the user demo.

s180
00:14:37.120 --> 00:14:39.839
The max amount is this, the currency is this.

s181
00:14:39.839 --> 00:14:42.080
If you want to revoke it, you can.

s182
00:14:42.080 --> 00:14:50.400
And what's the maximum Here, we didn't want to haggle, so we just put the maximum amount of that, and then it's also a single time usage.

s183
00:14:50.400 --> 00:14:58.000
This demo also has comes with a timeline, so you can actually open any of these and see these happening.

s184
00:14:58.000 --> 00:15:02.120
Remember that catalog I was talking about that they don't do the search?

s185
00:15:02.120 --> 00:15:06.360
We actually do a product feed, sending it to them.

s186
00:15:06.360 --> 00:15:16.080
Uh I have added that as well, and the feeds, because they're so different, there's a place to actually compare them.

s187
00:15:16.080 --> 00:15:18.240
So, here's the feed being called.

s188
00:15:18.240 --> 00:15:26.600
By the way, if you went to timeline, every couple of seconds, we try to get the catalog in sync for what is in inventory, what's not.

s189
00:15:26.600 --> 00:15:31.040
This is the UCP one, and there's the meta one.

s190
00:15:31.040 --> 00:15:37.440
So, I've shown you this, and you could reuse this same demo or the same concepts.

s191
00:15:37.440 --> 00:15:42.680
What if I didn't want to do external agentic commerce on Gemini or ChatGPT?

s192
00:15:42.680 --> 00:15:50.000
You could still build your own custom implementation of a merchant agent or Jenny on your website.

s193
00:15:50.000 --> 00:15:53.000
Maybe I start selling uh cat goods.

s194
00:15:53.000 --> 00:16:02.280
I could reuse some of this, but I would advise maybe look into some of these primitives and trying to use them because they've been standardized across merchants.

s195
00:16:02.280 --> 00:16:11.200
So, they have been well thought out, and also you could probably reuse them to sell externally as well on ChatGPT and Gemini.

s196
00:16:11.440 --> 00:16:15.240
So, remember I was talking about the discount codes?

s197
00:16:15.240 --> 00:16:16.440
There's a reason for that.

s198
00:16:16.440 --> 00:16:27.880
When we build out this demo and in my time building agentic commerce at as by, we've realized working with AI and conversational experiences without evals is playing whack-a-mole.

s199
00:16:27.880 --> 00:16:38.000
So, if you choose to use the same architecture for a new customer base, like jenny.com websites, think very much about creating evals.

s200
00:16:38.000 --> 00:16:44.839
Uh one of the things that I could not uh emphasize more about is you should test, test, and test.

s201
00:16:44.839 --> 00:16:50.480
If you go over here, I can also run my scripts for run evals.

s202
00:16:50.480 --> 00:16:55.400
And this evals folder has all this evals.

s203
00:16:55.400 --> 00:17:08.439
The reason I'm also showcasing the code is there's a template folder here, and we can go back to the slides.

s204
00:17:08.839 --> 00:17:12.959
And if you don't do evals, this is what might happen what might happen.

s205
00:17:12.959 --> 00:17:14.120
I love Chipotle.

s206
00:17:14.120 --> 00:17:18.680
I don't know if it's true or not, but I found it really funny, so I'm going to talk about this.

s207
00:17:18.680 --> 00:17:27.680
So, this popped up that when Chipotle rolled out their agent, people were using it to ask programming questions.

s208
00:17:27.680 --> 00:17:28.439
Right?

s209
00:17:28.439 --> 00:17:33.720
If you don't tell your agent to not allow for those kind of things, people will use it.

s210
00:17:33.720 --> 00:17:40.680
This is hands-down one of the most creative way to get free AI usage when you don't want to pay for that cloud subscription.

s211
00:17:40.680 --> 00:17:48.720
And if we don't write our emails and test intensely, those things will happen in production.

s212
00:17:48.720 --> 00:17:56.600
Uh the discount code will be told, even sometimes more uh uh sensitive things like who else is checking out this product.

s213
00:17:56.600 --> 00:18:06.360
So, the kinds of emails that I would highly recommend you write is behavior emails, protocol compliance, because when we're selling it to like, let's say, GPT, let's say chat.openai.com

s214
00:18:06.360 --> 00:18:12.600
or Gemini, you want to make sure that the feeds are actually me conforming, or else they will not support it.

s215
00:18:12.600 --> 00:18:14.920
You should also think about latency benchmarks.

s216
00:18:14.920 --> 00:18:26.760
Every second in retail on the shopping journey where you're actually not selling, there are chances that the other website's going to be faster, and people are just going to move away, or they just don't feel like it anymore.

s217
00:18:26.760 --> 00:18:31.200
Lastly, I also recommend using LLM as a quality judge.

s218
00:18:31.200 --> 00:18:33.080
Uh you don't have to use something fancy.

s219
00:18:33.080 --> 00:18:41.160
Talk to your product friends and figure out what's the best way to do it, and use a low like best use cases and write them out.

s220
00:18:41.160 --> 00:18:49.360
And I would like to also talk about Now, that I've discussed all of this, what's actually stable today and what's still forming?

s221
00:18:49.360 --> 00:18:55.160
MCP is widely adopted, A2A is widely is used, UCP ACP is out there.

s222
00:18:55.160 --> 00:19:05.680
Uh what's still forming though is AP2 and actual usage of it, ACP versus UCP convergence, do we always have to do two different specs, identity concept standards, and multi-agent

s223
00:19:05.680 --> 00:19:08.080
checkout delegation.

s224
00:19:08.120 --> 00:19:17.080
So, uh if you have to leave here uh today uh with anything, I hope you leave today with a good mental model of how an agenda commerce works.

s225
00:19:17.080 --> 00:19:19.800
I have nothing to sell you, but I do have gifts for you.

s226
00:19:19.800 --> 00:19:22.200
I find agenda commerce really exciting.

s227
00:19:22.200 --> 00:19:27.240
So, you can find this entire presentation on GitHub, and I came up with a template.

s228
00:19:27.240 --> 00:19:28.720
It's a three-service starter.

s229
00:19:28.720 --> 00:19:33.480
If you want to do custom agent customer agent or you want to do the merchant agent, you can do that.

s230
00:19:33.480 --> 00:19:36.520
Because I love e-vows and that saved my life.

s231
00:19:36.520 --> 00:19:38.960
I have some e-vow templates for you.

s232
00:19:38.960 --> 00:19:42.920
And you know, what if you want to send it to all of your merchants, not just one?

s233
00:19:42.920 --> 00:19:52.760
I have a catalog sync process as well, which will allow you to type into your own product and then turn it into ACP or UCP or meta, so you can sell there.

s234
00:19:52.760 --> 00:19:57.960
And lastly, but not the least, we all know now these days we don't write code like that.

s235
00:19:57.960 --> 00:19:59.920
If I give you a template, you'll be like, "Meh."

s236
00:19:59.920 --> 00:20:09.600
So, I have agent skills that specifically does a merchant agent customer agent and all these different catalog syncs that we have talked about.

s237
00:20:09.600 --> 00:20:15.400
I hope you had an amazing time and learned and had fun as much as I had presenting this.

s238
00:20:15.400 --> 00:20:15.840
Thank you.

s239
00:20:15.840 --> 00:20:20.680
My name My name is Priu, and I hope to see you again soon.

s240
00:20:34.034 --> 00:20:36.034
[music]