Exploring the Advancements and Challenges in AI Agent Deployment
John Nay, founder and Chief Executive Officer of Norm Ai, joins theCUBE's special presentation with NYSE Wired, focusing on the upcoming Artificial Intelligence Agent Conference 2025. Hosted by John Furrier, co-founder and co-Chief Executive Officer of SiliconANGLE Media, this insightful discussion covers the pivotal developments in AI infrastructure and the regulatory complexities faced by enterprises.
In this episode, Nay shares their expertise in regulatory AI infrastructure, particularly as it pertains to AI agent deployment in highly regulated sectors. The conversation, hosted by Furrier, delves into the evolving landscape of AI technology, compliance challenges, and the strategic initiatives underway at Norm Ai to address the pressing issues surrounding AI deployment. The discussion provides valuable insights for both technology and policy influencers.
Key takeaways from the discussion include the emphasis on the need for dynamic, real-time compliance frameworks that align with regulatory standards, as emphasized by Nay. Furthermore, the episode highlights how enterprises can leverage existing compliance structures to integrate AI technologies more effectively, offering a glimpse into the future of AI agent scalability and regulation. The conversation underscores the importance of bridging the gap between engineering, policy, and technology for sustainable AI innovation.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For theCUBE + NYSE Wired: AI Agent Conference
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for theCUBE + NYSE Wired: AI Agent Conference.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI Agent Conference
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI Agent Conference. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Eli Finkelshteyn, Constructor
Exploring the Advancements and Challenges in AI Agent Deployment
John Nay, founder and Chief Executive Officer of Norm Ai, joins theCUBE's special presentation with NYSE Wired, focusing on the upcoming Artificial Intelligence Agent Conference 2025. Hosted by John Furrier, co-founder and co-Chief Executive Officer of SiliconANGLE Media, this insightful discussion covers the pivotal developments in AI infrastructure and the regulatory complexities faced by enterprises.
In this episode, Nay shares their expertise in regulatory AI infrastructure, particularly as it pertains to AI agent deployment in highly regulated sectors. The conversation, hosted by Furrier, delves into the evolving landscape of AI technology, compliance challenges, and the strategic initiatives underway at Norm Ai to address the pressing issues surrounding AI deployment. The discussion provides valuable insights for both technology and policy influencers.
Key takeaways from the discussion include the emphasis on the need for dynamic, real-time compliance frameworks that align with regulatory standards, as emphasized by Nay. Furthermore, the episode highlights how enterprises can leverage existing compliance structures to integrate AI technologies more effectively, offering a glimpse into the future of AI agent scalability and regulation. The conversation underscores the importance of bridging the gap between engineering, policy, and technology for sustainable AI innovation.
In this interview from the Mixture of Experts AI Agent Conference, Eli Finkelshteyn, chief executive officer and co-founder of Constructor, joins theCUBE's John Furrier to discuss how AI is transforming e-commerce product discovery from keyword matching to intent-driven personalization. Finkelshteyn explains how Constructor serves major retailers — including Sephora, Under Armour and Petco — by providing an intelligence layer across search, browse and recommendations on their owned properties. He contrasts legacy keyword-matching engines, which often surface ...Read more
exploreKeep Exploring
What does Constructor do?add
Which metrics demonstrate how Constructor improves business efficiency (for example, conversion rate measured via A/B testing)?add
Can you tell me about your company—its history, technology focus, current size and growth, and the markets you serve?add
>> Welcome back to theCUBE Studio here at the New York Stock Exchange. This is Mixture of Experts, part of our program with NYSE Wired. This May, NYSE Wired is going on the road not too far right here in New York. We will be attending and covering the AI Agent Conference. Joining me now is one of the headline speakers from that day, Eli Finkelshteyn, CEO and co-founder of Constructor. Welcome, Eli.
Eli Finkelshteyn
>> Thank you so much for having me. I'm excited to be here.
John Furrier
>> So am I. The real premise of this conversation is we want a sneak peek for our viewers and audience around what it is that you're going to be sharing on stage in May. But ahead of that maybe, give me the 10 pence on exactly what Constructor is and what problem you guys set out to solve.
Eli Finkelshteyn
>> Sure, absolutely. And just before I start, I apologize for the background. I'm calling you folks from another conference, Shoptalk out in Las Vegas. I'm calling from a hotel room. But to answer the question in terms of what Constructor does, we work with some of the largest brands and retailers in the world, companies like Sephora, Under Armour, Petco. And what we do for them is when you go onto one of their owned properties, let's say their website or their app, you're searching for something, you're browsing for something, you're looking for something to buy. Our job is regardless of how you're looking for that thing to buy, we want to connect the right shopper to the right product at the right time in the right context. So making sure that the results that you see, they're personalized. There's something that's going to be interesting to you, there's something that hopefully you're going to want to buy.
John Furrier
>> So you have an interesting background that brought you to this. You were at Shutterstock for a while, I saw, along with your co-founder. I know a little bit about the Shutterstock platform myself and their big stock imagery platform, and how difficult it was sometimes to get the best possible return on just a human searching a prompt. It was a mixture of like NLP struggle meets just human struggle. And it does seem as though the AI era has certainly set out to solve that. Talk to me a little bit about what's changed from the perspective of technical opportunity though with unstructured data and NLP.
Eli Finkelshteyn
>> Yeah, I think a lot of things have changed. So back when we were doing it at Shutterstock, and this is the way the traditional search engines would work is they were primarily based on keyword matching algorithms. Not to go too deep into those, but basically it's like somebody searches for the word baseball and you look for other things that have that word baseball in them. The problem is that it will sometimes get spurious results, maybe a baseball mitt or something like that, which maybe isn't exactly what you're looking for. Or, it will miss some results, so maybe something that is very, very related to baseball, but doesn't actually have that term in it. So it'll focus on just keyword matching, but not really intent. What I think technology now has been able to create is that now we're much less bound by just those keywords. LLMs are much, much better at really understanding what a person means. And so that's created opportunity within the search space, within the discovery space to do something that's much less just technically relevant, but something that creates an experience that really matches the user's intent that connects them to the right product at the right time and the right context.
John Furrier
>> I mean, user experience is obviously so important. I'm a serial shopper myself, sadly. But also from the perspective of cost impact and commercial impact of having inventory that's not necessarily realized or the value is not realized because customers don't know how to find it, how do you quantify that problem? Are there numbers or metrics around that in terms of what Constructor can solve in way of making businesses more efficient?
Eli Finkelshteyn
>> Yeah. So usually we'll focus on something along the lines of conversion rate. So it'll be like essentially whether somebody actually buys something or not. One of the ways that we'll recommend to our customers to test Constructor before they buy is you send half the traffic to whatever you had beforehand, you send half the traffic to Constructor, and then you see what the conversion rate is on both sides. And if the conversion rate is higher within Constructor, then you kind of know that it's connecting more shoppers to products that they actually want to buy. And once you know that, then hopefully it becomes a no-brainer.
John Furrier
>> We've had some interesting conversations here on theCUBE of late around retail. We covered NRF this year. We had a lot of folks in while they were here in New York selling into the industry. And I would say there's definitely mixed sentiment around how impactful AI has been thus far in terms of really, I guess, revamping or reimagining the future of retail. I would have expected that it would be an industry that would have really taken off fast, but it doesn't seem to be everyone's understanding or narrative. What are your thoughts? Why do you think it is that retail has kind of lagged a little bit behind in the AI era? Or do you believe that?
Eli Finkelshteyn
>> I think it depends on the company. I think that there are many companies that are really on the forefront of it that are launching things like shopping agents, product insights agents. Just as two examples that are publicly known, Amazon and Walmart, they both launched. Amazon launched its Rufus AI shopping agent pretty early on. Walmart launched their Sparky agent. There are many others at this point. But you're right that most of them haven't yet. I think there's a lot of wait and see. I think there are a lot of companies that are trying to figure out where things are going that are maybe a little bit slower on the experimental side. But I do think that there's a number of companies, and these are probably going to be the winners in the future, I'd guess, that are much more forward on the experimentation of, now that this new technology is available, what can I do with it? How can I learn how to use it better faster than my peers? It's like with a lot of things that I think you have those that are on the forefront of it that are going to wind up winning.
John Furrier
>> Talk to me about the scenario whereby you bring on a new customer at Constructor. I, a long time ago, did some work in supply chain at Xbox, and I remember the thousands of spreadsheets, tracking bills and materials and different things, even just for one product. It was extremely arduous in terms of the overheads that's required. So you bring on a new customer. They have to in some way though document or tag their inventory, right? How does it work? How quickly and effectively is this plug and play? And how quickly can they actually realize value from the tool? Do you consider it a tool? Talk to me a little about what it actually means tactically.
Eli Finkelshteyn
>> We consider ourselves a platform just because you can use us in so many different places, whether it's your search bar, browse category pages, recommendations, offsite and email recommendations for AI shopping agents, anywhere where you can discover a product. We think of the platform primarily as providing the intelligence layer; so given the context, which product is the right product to connect the user to. The way that you install it though, it's relatively easy. Most companies will already have a product catalog somewhere. Maybe it's in their e-commerce platform, maybe it's in their product information management system, maybe it's somewhere else. All that they need to do is send us that product catalog in whatever format is easiest for them. At this point, we've gotten pretty good understanding just about every format. We've got a connector for just about every format. There's even one that's just general where you just send us whatever and we'll map it ourselves. And then the other only thing that you're doing is you're connecting us to clickstream data, so what people are clicking on, what they're adding to their cart, what they're purchasing in an anonymized way. So we can start learning based off of that clickstream data which products are getting the most clicks, the most add to carts, the most purchases, which ones are getting scrolled right past. It's really understanding from your shoppers which of your products are most interesting to them in a given context, which them are not interesting to them. And then based off of that, figuring out what is the right product to show to the right customer at the right time and the right context.
John Furrier
>> Let's talk for a second about Amazon and that model, because it is an interesting time from the perspective of retailers doing direct to consumer and using the kind of wholesale model through companies like Amazon. We've heard a lot about what it means from the perspective of your margin when you use companies like Amazon, but they offer a lot of advantage from the perspective of speed and reverse logistics, et cetera. We also know that this LLM future could fundamentally shift the whole world of search engines and how people think about what results and return they get. How do you think about this future from the perspective of the companies you're working with? Do they want to do more and more of their own in house, direct to consumer, have customers shopping directly on their sites? Is that like the ideal future for the AI era? Or what are your thoughts?
Eli Finkelshteyn
>> I think that companies generally are afraid of disintermediation and for very, very good reason. If they have the shopper relationship, then there's so much that they can do with it. There's brand loyalty. There's the idea that people will keep coming back directly to them. But if there's something that stands in between, whether it's an Amazon, which by the way, I'm a big fan of, but it does disintermediate; or it's something newer like Google where now they want to have shopping done directly through it, what it means is that you get disintermediated from your customers. So now when a customer goes onto, let's say, a Google and they're searching for something, if the purchase happens directly through Google, or through Amazon in your example, then there's no guarantee that that now service that stands in between is going to care about your brand loyalty. Somebody goes on there and maybe they search for a shirt. Even if previously they've purchased many Under Armour shirts, if they're searching via Google, maybe one of Under Armour's competitors might decide they're going to bid more on that query. Maybe for whatever reason, Google might decide to show more of that competitor's products. And so now instead of having that brand loyalty that you're used to and you're hoping to create, you're constantly getting challenged on it. So the safer thing, and for what it's worth, I agree with brands and retailers on this, I think that it makes sense, the safer thing is to get people to want to come via your own channels. The own channels are the place where ideally you make some experience that the user really enjoys, whether it's currently on a website or an app or an app, maybe in the future it's something like via ChatGPT app, but you control more of that experience. You're getting people to come to you because they specifically want to come to you. And you know that if they're coming to you, they're not going to buy from somebody else, they're going to buy specifically from you.
John Furrier
>> And I guess loyalty programs and things too, right? They have been super effective as well. So you are going to be at the AI Agent Conference here in New York and May because Simon, John, and that team see you guys as a very big disruptor for the future of retail. Talk to me a little bit about the company itself in terms of where you're at. You've raised a number of rounds. Give me the commercials here. Fill me in.
Eli Finkelshteyn
>> Yeah, we're about 10 years old at this point. We spent the first four years of the company building. We're a little bit atypical in that. What we wanted to do was build the whole system from scratch specifically for e-commerce and specifically on a foundation of AI. We thought at the beginning, I guess before it was as popular as it is now, that the right way to solve discovery for e-commerce is specifically based on that clickstream that I was telling you about, really listening to users, understanding based off of their clickstream, which products are interesting to them, which ones aren't. And then we spent the last, now it's about seven years actually, in market. At this point, the company is, I guess, roughly 400 people. We're still, in terms of customer growth, growing quite quickly. We announced that we were growing, in our last momentum release, it was about 80% growth in terms of customers over the last year. So we're pretty excited about the future. We're excited about what we've done so far, but we're also really, really excited about the future because I think there's still a lot of companies within retail, within both B2C and B2B brands, manufacturers that we can help, help them connect the right product to the right shopper at the right time and the right context.
John Furrier
>> For sure. And you're talking up some pretty nice headline, the accolades there too with Fast Company and the Gartner , et cetera. So congrats to you and the team. What can we expect to hear in May? Have you got any sneak preview in terms of what you're going to be chatting about on stage? Are you joining a panel? Give us the .
Eli Finkelshteyn
>> I'm not sure yet in terms of the format, but it's the Agentic List Conference, so I assume I'll be talking a little bit about some of the agentic things that we're doing. We actually just had a new release at the conference that I'm currently at, Shoptalk, of our merchant insights agent, which we expect to be able to really help merchandisers, business users, the folks that are in the platform be able to perform much faster. But I'm also really excited about the shopper facing things, so our AI shopping agent, our product insights agent, talking about some of the success some of our customers have seen with those things. I'm excited about traditional forms of discovery, of course, but I think that some of the agentic stuff can fundamentally change the game and I'm excited to talk about some of that stuff at the conference.
John Furrier
>> Well, we're excited to hear. I know you're dialing in from Vegas. What time is in Vegas right now? Early in the morning. I'm sure you need that sugar rush to get the day going. So listen, Eli, thanks so much for joining us on theCUBE and NYSE Wired. I look forward to seeing you in May.
Eli Finkelshteyn
>> Thank you so much for having me.
John Furrier
>> I'm Gemma Allen. This is Mixture of Experts, part of our program with NYSE Wired here from theCUBE Studio at the New York Stock Exchange. This day, we're talking about the AI Agent Conference happening in May here in New York. Thanks so much for watching.
John Furrier