 Welcome back to theCUBE's coverage here on the location at ReInvent Amazons. Web Services annual user crisis. theCUBE's 11th year. I'm John Furrier, your host with Dave Vellante, extracting the similar. It's been quite a journey, Dave, and been doing this for 13 years of coverage, 11 years of ReInvent. Love to get to practice this long. Got a great guest, Michael Schultz, with us, vice president of product and customer market at Commerce Tools. Michael, welcome to theCUBE. Thank you, thank you for having me. You know, one of the best parts coming to ReInvent is besides swimming through the crowd, which has gotten worse. Say to the right, I mean, it's so crowded here, but the practitioners are here. You got real builders, it's an educational conference, as Aidy Jassy always used to say, and now Adam Slepsky, they point that out right out of the gate of the keynote. A lot of stuff going on in this year with Jenervaeye, and yeah, they leave the keynote up with storage. Dave and I are talking, and he's like, come on, isn't this exciting? It's storage, I mean. And I love storage. We love storage. Snorage, there's storage though, storage is hot. Infrastructure's hot, obviously the Jenervaeye stack. You know, infrastructure's going to be a big enabler. Huge opportunity. What's your take of how this is all coming together? I mean, we think we're in a great position. So Commerce Tools in and of itself, like, we are a composable commerce business. So we give all of our companies and customers the opportunity to pick and choose all the components that they need to build outstanding shopping and buying experiences. And we do that with a cloud native platform. So we're built in AWS, we're built in GCP, we have a multi-cloud infrastructure strategy. And so we're built in the cloud for the cloud. So and we actually think that being API first, being cloud native is the best foundation for Jenaeye or any sort of AI ML use cases because we expose all the data, like that's hence why we use MongoDB as well. There's nothing sort of proprietary or lockdown or a black box. And so we think we're at a good point from a company perspective to really, really do this well because we're exposed to everything. You're essentially a headless platform for retail, right? I can, as a retail operation, use your platform and customize it for my specific needs, right? Yeah, that's exactly right. So whatever your customers demand from you, whatever customer experiences that you want to do and also across different verticals, business models, so B2B, B2C, any hybrid models, marketplaces, subscriptions, we can expose all of that functionality in a true cloud native API first fashion. And so you can build your experiences the way you want. So Michael, talk about the impact of Code Whisper. We saw some great demos up there of how they're using the data with Jenaeye, Code Whisper for developers, and obviously the business users got agents. So I can imagine that this is going to be a real enabler for no code, low code. 100%. Your customers, if you have that system behind it, you have to take advantage of that to enable that. How does that play out? If you take us through your mindset when you hear the announcements of this new three layer stack from AWS and you get Mongo as your partner, what's the opportunity for your customers to do with this new Jenaeye? I think the opportunity is tremendous. I think we're optimistic, but we're also pragmatic. So the way that I look at it is in a couple of ways. Number one is how can we enable business users to do their job best? So one example is promotion, personalization. It's even maintaining product information, like really accurate and do it well. I mean, Jenaeye is good for creating content. And if you think about changing product mixes, adding products, acquiring brands, like we want to support all that in a way that we can help our business practitioners to create these amazing and stunning customer experiences because they have the most and up to date like content available and the product descriptions available. And I think there's more. So there's personalization, there's promotions, there's product information. I think very, very key to us. Since you mentioned low code, no code, I think we want to be in a position from a technical side of things where we literally want to go to chat GBT or Jasper or your tool of choice and say, build me an integration into this promotion engine, into this marketplace vendor. And we effectively give you a prompt library to do that. Like we literally took the entire adidas.com website and we wrote a prompt library of how can you build adidas.com in six hours. We wrote a block about it. And so I think that's where the future is. So it's low code, no code. It's kind of like, how can you empower like- Can we get some of that? And then the quality of that code, how much human interaction posts all those prompts does a person have to do to get it to where you want it to be? Is it halfway, three quarters of the way, 100% of the way there? I think we're still very early. I think the benefit that we have is all of our API documentation is exposed. It's beautifully written. I mean, it's technical and like, but you have everything open source. It's very transparent. So as long as that is like the foundation, if that's transparent, if that's open, if that's accurate, I think there's very little like post work required. But again, I also think we're still very, very early in training these models. You have all these large language models out there. Like it's almost like commercialized in the future. So I think you got to really figure out who's the vendor that you want to go to market with? How much do you believe and trust and honor that quality from that LLM model? And I think it's there. I think we're really empowering kind of like the business users, but also the tech users and our partners as well. So it dramatically cuts the time that it takes to research and get to an answer, right? I mean, it's all about improving efficiencies like hands down, everything that we're doing commerce tools, everything that Gen AI is doing. It's all about improving efficiencies. And I think if you look at the macro environment, whether it's geopolitical conflicts, supply-gen issues, inflation, like retailers, I think in a way have the hardest job at the moment because they have to figure out where's my demand coming from? We just finished Black Friday, Cyber Monday, right? So I think planning for the demand, catering to that demand is super key. How were the Black Friday and Cyber Monday trends? I was driving down the Cape the other day, John, and there's this retail outlet, the rent them outlets, the cheaper outlets. There was two mile backups on each side trying to get in. I'm like, okay, so the economy's doing okay, the consumer's spending, what did you guys see? What we saw is, I mean, we saw the physical component growing, but certainly not as much. I think we saw maybe like a five, seven percent uplift on physical store sales. Online, we saw it skyrocketing, 30, 40 percent, yes. And so I think what we're trying to do is we're trying to cater for this sort of really omnichannel exposure that retailers have. It's no longer about pure play online, it's not about pick up and store, it's all of the above point of sale, client-telling, virtual try-ons. Like how do you capture data? How do you cater to that? And how can you actually transact in the moment? That's really important for us. Do your customers, are you bundling in, for instance, the AWS infrastructure cost, or do your customers sort of pay for that separately? How does that work? So we do follow a broader hybrid or multi-cloud strategy. Okay. On some marketplaces we are available, on some marketplaces we will be available in the future. I think it's always interesting to look at us as a way that we're unlocking the entire AWS ecosystem. So whether it's certain vision APIs, big data like aspects, like we unlock all of that. Again, thanks to our platform and MongoDB because all the data is there. It's not like buried or in a black box somewhere in like a monolithic legacy platform. All the data is exposed via our APIs. So that effectively can be pumped into any sort of AWS ecosystem partner and they can use that data. So you got the transactional app on top of MongoDB. And then how have you begun to use generative AI? Are you using vector database? Are you using Mongo's capabilities? Have you started playing around with Mongo's vector search? Yeah, so I think last year I think we really started using MongoDB search and playing around with that. I think search is so critical to the retail aspect, everything around discovery and engagement, right? I think B2B companies want to be like B2C companies. So I think B2B was usually about finding things and B2C was more about discovering things or being served things like really in a personalized way. I think what we're seeing is B2B coming closer to the B2C realm at the moment. And so we're trying to figure out like, how can we support all these different nuances around discovery, engagement, tracking the entire customer journey. So we're looking into the AWS kind of like ecosystem. We're looking into MongoDB, everything that they're bringing out. I think for us it's really important to work with a partner like MongoDB because they're exposing all that gen AI stuff to us and we can pick and choose what we want to sort of incorporate in our platform and then surface to our customers. Take us, first of all, yeah, this whole agent stuff's going to be boom for you, your customers. They're going to love that automation, doing some cool retail stuff. Explain your relationship with Mongo. They're a partner of yours. What are you doing working with Mongo for? What specifically is the products that you're using and what problems you're solving with them? Yeah, sure. So MongoDB effectively underpins everything that we do. So again, we're hosted on GCP. We're hosted on AWS. Everything is working and on top sits MongoDB. So it's effectively the data layer, the data platform that powers our platform. So we're using MongoDB Atlas and it's effectively underpinning and powering everything that we do. And for us, I mean, in the bare bone sense, we are a transactional engine capturing every single digital and physical transaction for our customers. So for us, it's important to use products like search and sort of vector to figure out how can we better serve our customers so they can serve their customers better. And so I think that's key. I think everything around the AI extensions, real-time data, kind of like really being able to like edit, capture, translate and rich data and then serve that up back to our customers is super important. Are you tapping services across those respective clouds or do you stay sort of typically within a single cloud? We're completely agnostic, like whatever you choose. But so if I deploy a platform, a commerce platform, am I sometimes, can I pull in? Can I use, say, OpenAI running on Azure? Can I use for the same instance of? You could, I mean, at the end of the day. Are people doing that? I think, yes, is the short answer. I think it really depends on the use case. I think a lot of companies have, like they're either a GCP shop and AWS shop, for us, it doesn't really matter. We cater to all of these scenarios. We have a lot of customers that have, effectively, they're chosen to a lot of pods and they're trying to synchronize that. At the end of the day, the net net is that, given the way that we are architected and the way that we're exposing data and content and transactions through our APIs, you can use it in any sort of connection and system that you choose to. Because one of the big themes we heard in the keynote today was LLM optionality. I might want to use, you know, AnthropaQ here or Titan or maybe I'd like to use OpenAI or VertexAI. You're saying, in theory anyway, you can enable that. At the end of the day, like a cloud native approach and an API first approach can enable a lot of that. And that's what commerce tools is built upon. So I think that that's possible. I think the lowest hanging fruits at the moment are really focused around product recommendations, personalization, and product information management. Those are the lowest hanging fruits, very low risk, very high reward. And then I think whatever is in the future, whether it's virtual try-ons or sort of AI VR, kind of like aspects or opportunities, I think that's where AI can come in as well. And to deliver those capabilities today, you're saying keep it simple, keep it low cost, make it fast, it's in ROI. I mean, again, I don't want to, I want to be realistic and this is why I'm saying, we're very pragmatic about this. In a way, everything that we're hearing about gen AI at the moment, sounds like big data 10, 15 years ago. So we are very pragmatic, just because you're a vendor and you put AI at the end of your product, doesn't mean you're really AI enabled or have that capability. So we encourage our customers to see what we have under the hood, and also what can you do in the future? Keep your options open. I mean, Michael, that's such a great point. Being pragmatic doesn't mean you're slowing down or you're decelerating, as they say. You're in a deceler and a cell. The key is the data. Like what we're seeing in our reporting, and you see what today, I think this is why I like how you have this whole data layer. If you enable the data to be available, that's a setup for the AI. 100%. So yeah, low hanging fruit, some wrapper wraps here, maybe do some stuff around agents. What do we have for data? And then I think the gun, the next question is, okay, what are you enabling your customer to do? Are they developers? What kind of developers are they? Are they app developers? Do you want to be business developers? So I think that whole discussion of what you do to enable that utility for the customer, is it the data layer? What in the data do you see you enabling your customers? Because I can see, makes logical sense. Mongo is the layer across multiple environments, but now you got to think, okay, what's the customer going to do? I think I'm going to write code. Is that code whisperer? You guys are going to have your own thing. So how do you see that data management layer philosophy? What's that? I think my philosophy across my career was, like if the data is not good, if there's no data hygiene, you're doomed to fail. Every project that we're starting, we're starting out with what does your data model need to look like to be successful? Because you can have all the big data, you can have all the in-memory data. At the end of the day, you just have bad data faster, and that doesn't help anyone. So for us, it's really have full, complete access to data, and make sure that you can slice and dice the data, move it around, translate and reach things of that nature. And so what you choose to do with that data is completely up to our customers, and we encourage to do so. But at the end of the day, for us, the single source of truth is in Mongo. How has the data hygiene equation changed, or how will it change as generative AI gets more steam, where prompts are actually embedded in stuff, and the apps are getting smarter? How do you see that hygiene evolving? Certainly it's not going to go away. It's not going to go away, but I think it's going to be improving. I mean, I look at a future where you're working with like big multinational brands like LVMH, right? They have 180 maisons, everything from Christiane to your what-have-you. I'm looking at a world where you say, here Mr. Chad GBT-4, I'm going to give you this product catalog, and here's my brand and style guidelines, and I'm just going to throw it at you, and you write product description beautifully designed in the brand voice and tone, and do this across my entire portfolio. And obviously, the data needs to be there, the data needs to be exposed, and I think that's a future that I want to see where, again, it's probably not going to be ever 100% perfect, but you're increasing the dupes, the misspellings, things of that nature. I mean, and you're better targeting the customer requirement, that personalization that you're talking about before, because shoppers often get very frustrated because they can't really find what they're looking for, they saw something one day and they can't find it again, so you would think that fidelity would be dramatically improved, and my question is, do you see a future where that becomes more real-time in nature, with more intelligent apps, essentially, on top of the platform? I think so, and I hope so, because again, talking about the search, discovery, engagement part, I think having that data available, and again, following the entire spectrum of the customer journey is key to capture that data and service your customers better with the data that you have, so it's not all about personalization, but you will find an increase in uptake there. Michael, great to have you all on, thanks for your sharing, love the data hygiene, love the vision, clean data, I mean, you got to have good data. Good data is good AI, is what we always say. Thanks for coming on, appreciate it. Thanks so much, appreciate it. I'm John Furrier, Dave Vellante, we're going to send it back to the studio for more coverage coming back here. Stay with us for more exclusive coverage from Reinvent, Amazon's user conference, is theCUBE, be right back.