 Hello and welcome to this special CUBE AI Conversation. I'm John Furrier, host of theCUBE here in our Palo Alto studios with a great guest here, VJ Chaturth, co-founder and CEO of BlueShift. You guys have been ahead of the curve on AI. Welcome to theCUBE, I mean AI is the hottest thing. We just both came back from re-invent. Congratulations on your success with AWS. Welcome to theCUBE. Thank you, thank you for having me here. Excited to be back from reinventing and talking about AI. We had your co-founder on as part of the startup showcase. So much has happened since then. But you guys have been ahead of the curve on this data developer concept, the 10x data engineer, data engineering, your history and pedigree in tech has been well-documented. You guys have been on that wave, personalization, scale, data. Now, everybody wants to do general AI. It's the hottest thing and certainly AWS completely talking about it with the three layer stack. And it's fundamentally changing every single company and every single industry and vertical up and down the stack because data is now the competitive advantage. Yep, absolutely. This is like the number one conversation and the impact is going to be profound as a generational shift kind of like we saw at the web. We saw it mobile and now AI. So a huge conversation. Can you scope in your mind the opportunity and the problem statement that you guys are attacking? You guys have an interesting approach. You're right in the wheelhouse of the data engine of what's going on and it's feeding the AI. What's the problem statement that you guys are trying to solve? That's an opportunity. That's a great question. So really we positioned Blue Shift as an intelligent customer engagement platform and the problem that we see is that every brand today suffers from this idea of siloed engagement across the customer journey. So if you and I are consumers of many, many brands and in our own personal experience, we see that our experience when we walk into a bank branch might be very different from how we get engaged inside the digital product or how we get engaged when we call the call center. And really this problem, if you think about it, people are trying to solve it. The traditional way of solving it is by building a bunch of campaign decisioning rules on how to engage customers. And often these rules look like if this, then that. But today the complexity of data, the amount of data customers are leaving behind the number of channels on which they're engaged. If this, then that paradigm completely breaks. And that's leaving this huge, huge problem in front of all the brands as to how do they solve the engagement silo problem and get to intelligent customer engagement for the new world. And the way to solve that has to be about moving away from this manual decisioning paradigm to an AI-driven decisioning paradigm. That AI-driven decision paradigm has to be supported by unified data across the entire customer journey. And it has to influence decisions on all the different engagement channels. So in some ways, when we think with the solution and the opportunity, it's about the 3Ds, the data delivery, the data decisioning and delivery. On the data front, it has to be able to unify data in real time. On the decisioning side, it has to be able to moving from manual decisioning to AI-driven decisioning. And that's the part we'll talk about a lot when we talk about AI. And finally on the delivery front, being able to make sure that the decision that your AI comes up with on how to engage, what to see to them, what offer to put in front of them, all of that can be conveyed on every different channel of engagement as those channels proliferate today. VJ, talk about the customer point in time. Where are they right now? Because unifying data could be easy or super hard depending upon what data you're trying to unify and what stack you're going to move across. And the automation piece, again, assumes pre-existing data. And also the consumer behavior's changing. A lot more organic third-party, the first-party data's changing. Where's the customer right now, as they're on the cusp of this next generation of AI? Are they at the beginning? Are they kind of moving down the line? Do they have enough data to unify? And how hard is it to unify? I know it's kind of a big question, but where are they right now on the journey? I think everyone has tons of data. We do, every enterprise we talk to has tons and tons of data, which is being generated by end consumers like you and I interacting with them on many, many more touch points. The problem that everyone faces is, how do you activate that data? Because a lot of that data is latent. It's not influencing the end customer experience. So to make that data actionable, to activate that data, you need to unify it across many sources and unify it in form, which is accessible to the end engagement application. And oftentimes, I think you asked about where are the enterprise, where are the customers in this journey? I think what we are seeing is that many of them have started a journey to at least bring different types of data together in a cloud data warehouse. That is an important step. It needs to be complemented with combining that with true real-time data because enterprise data warehouse are great for many things, but at the same time, if I expect a response in a second, if I am on your website, that's not an application for which you want to query the data warehouse. So how do you combine this real-time event data with the relational data that's in data warehouses is one of the key challenges that needs to be solved. And once you solve that and you truly unify the profiles in real-time, you need to be able to make AI decisioning on in real-time as well. So I would say like every brand, every customer that we talk to has taken the first one or two steps, but the journey ahead of them is really about unifying that with real-time data, making AI decisioning in real-time and distributing that to every different channel. It's almost like just when they thought you unified the data, more data comes in that needs to be unified. It's like a circle that keeps its flywheel there. Question I want to ask you is, can you give an example of what it means to unify your data? What does that mean from a customer standpoint? Can you walk through an example? Yeah, so I think like instead of unifying data, let's think about unifying the engagement. So as we mentioned, the problem is if you don't do that, there are real examples either today or even from when we founded the company, we saw some of these problems where a single customer might touch the brand on different touch points and end up getting 14 different triggered emails, SMSs, notifications on so many different ways which are all saying different things. So the engagement, broken engagement problem is real. Obviously it is, the reason it exists is because of fragmentation of data, but I think I would encourage brands to think about solving that eventual problem of unifying the engagement. So step one is solving the data which is making sure if I interact with a digital product which might be a logged in experience versus if I'm browsing the marketing website which might be an anonymous experience, you're able to unify that. If I walk into a store or a bank branch versus if I'm on digital, you're able to kind of connect those pieces of data and you're able to connect the contact center data with that. And to do that, you have to solve obviously some of the identity problems. You have to solve this idea of creating an index of that data. You talked about search, for example. If you think about the reason why Google search is blazing fast is in that moment they're not going in literally reading all the pages. They have done some bit of that. But they also make that index living and breathing with new content that's coming in real time. So really finding ways of integrating data where you can absorb real time data but you also have an index longitudinal view of the customer starts becoming very important and solving that engagement silo problem. So they had a lot of data laying around. I see that in a positive way. It means this data to be leveraged. The classic data exhaust is data goal deal expression. Can you give an example of insights in the automation and explain that, how brands are leveraging the AI specifically? What's the low hanging fruit and what's kind of like an advanced version of use case? I can see them seeing, getting the data, quick wins out of the gate and then what happens next? Cause you can do a wrapper here, you can throw a search on it and then reasoning is a big promise. We're seeing kind of three levels, small, medium and large use cases to get us through that. Yeah. So I think when you think about the use case evolution, we talked about the unified data but then the second step to solving the engagement silo problem is AI decisioning. And when we think about AI decisioning, we think about sort of figuring out who, what, when and where kind of decisions. So who is really about which customers to target in this moment with any particular offer or message or content. And if you think about it, we have a company customer in the financial services space. They are on an ongoing basis trying to predict whether this customer based on combination of demographic data, behavioral data, transactional data are they in the right frame of mind to receive communication about mortgages versus credit cards, for example. And once you meet that determination, you got to get more specific. So that's the who, but you got to figure out what do you put in front of them? Is that a piece of content? Is that an offer? Is that a product? And that could be in this example a specific credit card offer, for example. And then you got to then decide the when and where. The when is all about the stage of the life cycle journey when you want to engage with that customer. It might also be the time of day because as you know, we're all busy and we are at least in the middle of this interview we are not doing this, but often we are meeting and we are checking emails or messages but we don't intend to act on that in that moment. So the finding the right time when the customer would engage becomes important and then the where part of it is very important which channel of engagement should you use? And there are paid channels like media for a paid media, for example, or expensive channels like direct mail. Should you deploy those or should you use more organic or own channels? Those decisions become important. So really I think decisioning based on that data the use cases are all about decisioning on who, what, when and where. I think that's maybe a starting point. But then coming to this notion of gen AI I think everything I talked about so far is what I would classify as customer AI. So how do you combine customer AI with gen AI to drive personalization? That I feel is the exciting opportunity ahead in front of every brand and that's where we think the market is going and that's where we are investing a lot to get to that next level of personalization. I love that gen AI and customer AI because it implies the customer has data that's proprietary or intellectual property. Proprietary not a good word but it's still their proprietary data not like proprietary software but. And then gen AI there but I want to just get back before I get to that. You mentioned the channels. I mean the efficiency around costs, right? If you can use AI to figure out that hey if my most expensive channel to reach the customer I can reduce the risk of waste or leakage. I can maximize the payload. That's exactly right. For that channel, that's a cost benefit. That's a massive cost benefit and also it's one of those things which is a great it's a better customer experience and it's great for the brand because it saves money. I remember when we started the company prior to that you know I founded a company that became Groupon Goods and then I was at an AI company which became Walmart Labs and companies like Walmart and Groupon we saw this problem that we would be spending paid media dollars against customers who might have converted any which ways and we had no way of like figuring out how we make those decisions. So obviously unifying the data is a step to that but the AI decisioning is absolutely important. And again, your pedigree and your guys background is so right on for this because the personalization is a benefit to the customer. That's right, yeah. And then the cost performance of the channels that's a business benefit. Total win-win, great example of good use of machine learning AI and good stuff. Now let's get the generative AI and customer AI. You mentioned that. So generative AI, the word generative means it generates stuff. That's like you see streaming words and an answer. Now they got multimodal, we got a little multimodal thing going on, there's a video from this, that perplexity, it's got a little thing in there. Really hard to do but it's cool. But it's got hallucinations, it's from an open closed web I guess now. But everyone sees that benefit. What came out of re-invent is interesting because you're on this. It's what you do with your own data mixed in almost the alchemy of models. That's exactly. Integration of models, APIs, how they connect, how to manage that, how to check against the LLMs is a quality. This is the new science. That's exactly right. It's playing the customer AI aspect, how generative AI is spawning new categories like customers AI. Yeah, so I think when you think about generative AI to oversimplify it a little bit, it helps make the content creation process a little bit faster. And traditionally, if you think about the world of customer engagement and marketing, there has been a content bottleneck. Like people would say, well, I need to run this engagement campaign, I need to deliver personization. But to create the content would take like four weeks. And now suddenly you can imagine that in some ways, I mean it's not literally true but often that four-week window can be shrunk down to minutes or obviously there are a couple more steps which I'm glossing over. But you can imagine that the time to create content, time to create variations of content is shrinking dramatically, right? So now there's like explosion of content. But explosion of content by itself, unfortunately it's a spam as we all know. But really I think how you solve that problem, explosion of content and make it relevant is by combining that with by matching the right content or the right content variation with the right customer. So how do you do that? You've got to understand each and every customer. You've got to understand using customer AI, what their preferences are, what their affinities are, what their intents are. And then be able to match that with the right variation of content that would appeal to them. Simplistically it might be about creating, when you create a SMS message or email engagement, creating variations that appeal to Gen Z versus a different segment for example, or it might be more cheerful versus casual or persuasive. But you can imagine it going much more beyond that and getting to an individual level, where we understand, start understanding individuals through the lens of the type of content that they are gravitating towards and use that customer AI to then match them with the right generative AI produced content. I think you're hitting on something that's going to be a big conversation over the next couple of years. It's already kind of coming out now and the industry is starting to see people say, hey, Google SEO is getting killed because the SEO is being gained by content. Okay, which is changing the SEO. By the way, it's about time. People don't really use Google search, but people are now going back to the idea of going to a destination because it's trusted. And this idea of vanilla content or generic AI content is not going to attract a lot of targeted audience. It's going to be like broad, people ignore it. But injecting customer data into that customizes the generic content. This is where I think you guys are onto something. So explain again how a customer would take their data set and go to say I'm producing a content farm of AI, hey, you know, whatever, X, Y, Z. I then can vector in customer content, my content, to give that an accent or flavor or spice or give it some sort of lift. That's exactly right, yeah. That's exactly right. So when you think about this idea of customer AI, you can understand customers through the customer journey path they are on in the past, without generic AI, you would just think about them, you know, in terms of funnel in some ways or their behaviors or their transactions or their demographics. And I think what this idea of LLMs and just the world of language has created is starting to understand customers as they interact with the world of language, right? And you start thinking about the embeddings of the content that they are interacting with and being able to layer a representation of that onto the customer profile and being able to use that sort of knowledge to be able to query back the right variation of content that would appeal to the person to drive their personalization. That is kind of what we're talking about and that's super exciting. Yeah, I think you're on a big wave. Our CUBE research team, formerly Wikibon, we just renamed it, is doing a big research study on AI and what area they're doing is, and I want to get your reaction to this, is looking at, okay, if brands are telling us this, okay, we want to tell our own story. We want to go direct. Okay, great, they have a website. Website's old, they got mobile apps, web apps, whatever. And so they're trying to figure out their first-party apparatus for the data. And then they're pushing content into these walled gardens like LinkedIn and trying to get people to pull back into their system. So thinking of it almost like a, they go out to the crowds, they push the content out where they are and they hope for some sort of round trip. Not all outlets have measurement, Twitter's got some, others don't, and there's some measurement around there. How do you talk to that brand and say, okay, I'm going to unify my data. I want to remodernize my first-party system right now. What's that look like? Talk me through that. I want to tell my own story, but I'm going to go publish in these venues that might not give me any analytics or leak analytics. That's exactly right. So I think going back to, I think you'd use the word maybe flywheels right in the past, and I think this idea that, you know, traditionally everyone thinks about marketing and customer acquisition or customer attention as a funnel. You start with like advertising on, you know, Twitter or whatever in your example and like kind of, you know, assuming for some percentage of that, those people to come back and there's a funnel. And I think now the modern flywheel is really about brands trying to say, well, I delivered, you know, a piece of engagement. It was relevant and that created little bit of data. I activated that data and delivered an even more relevant experience back and that pulls customers back. And then as they do that, they create more and more data. So a lot of that data is going to be captured. I would say outside of the wall garden, maybe on your own properties to a large extent. So before talking about data and infrastructure, you want to kind of talk about the key to it all is about creating those trusted engagements with the customers, right? Like being seen as their trusted destination. I think you talked about some brands creating those destinations. So I think having the right trust, having the right personization which brings customers back is key. And once they do that, the data infrastructure has to be about sort of, you know, being able to understand data from every part of the customer journey. Some of that data would be in CRM systems if there are human interactions kind of happening. Some of that data would be event stream data, which if it is digital interactions that are happening and you got to be able to combine, you know, these digital offline interactions into a single customer profile, you got to be able to look at different identities of the same individual. Some of them might be a logged in experience, some of them might be anonymous. So how do you do all of that and, you know, unify that into a customer data platform, be able to use AI. So all of the technology decisions become much simpler if you think about first customer objective, which is about, you know, creating those compelling experiences, creating that flywheel in the first place. Okay, so how do I get started? How do I get this going? I'm sold. Say I'm a customer, I can sold me on at VJ, I'm in, what do I do? How do I deploy? How do I engage with you guys? Take us through the engagement process. Yeah, so firstly, exciting announcement. We are making that deployment experience dramatically simple. So what we are doing today is launching a free experience where you can sign up, you can start unifying your data warehouse data with real-time event stream data. You can segment your audiences without even knowing SQL or anything technical. You can, you know, connect that to a multitude of destinations, including advertising destinations, marketing destinations, and customer experience. And you can orchestrate both audience engagement activation, as well as one-to-one, right? Which is what we are all trying to get to. And all of that is available now for free if you go into BlueShift and sign up. So it's a great way to start your journey. How many steps does it take to get going? So we feel and we have tested this out with many customers in ourselves. Within 15 minutes, you should be able to get a lot of compelling value. And that compelling value can be in two forms. Within 15 minutes, you should be able to take data from multiple sources, but be able to orchestrate a journey across two or three different touch points. For example, if you're in the hotel booking business, you could orchestrate a journey which, you know, starts all the way from, you know, prior to the booking is made to looking at times, you know, when the booking is about to come up and post-booking, like, create an NPS-connected engagement and be able to connect different touch points. And all of that you could do literally within 15 minutes. You could pull in data from a source like Snowflake. You could, you know, pipe in, you know, a couple of, like, events. And you could configure, you know, these engagements to happen over an NPS application, over an SMS application, over an email application, all in one journey and it's very, very quick. So this is targeted for the brands who don't want to take the big risk of doing a POC or might be kind of scared, looking at the tech stack and saying, oh my God, if I implement this, it's going to have disruption to my pre-existing martech stack, people who just want to get, try it. Yeah. It's a starter kit. It's a great way to, it's exactly a starter kit. It's a great way to get started because you talked about the journey of every brand. I think, like, it's a big unlock in their mindset saying that all of this is actually now possible. It's possible within, like, a matter of minutes with a, you know, with a team of one person or a very small team. And I think that's the key to today's modern engagement if you think about sort of democratizing access to all this technology. Because back in the day, you know, doing all this would have taken, like, an army of engineers and data scientists and now we're saying it's all super simple and I think that's the key to unlocking many different animals. That's available today. It is available today, absolutely. Where does it go next? What's your vision as AI continues to grow? Reminds me of the web. It's very embryonic. You're going to see instant wins, ease of use and simplicity. There's always been a key factor in inflection points. You guys have been there, done that. Yep, that's exactly right. So I think our belief is that, you know, AI technology is one of those foundational technologies because, you know, back in the day it was, it required, like, technical people to understand and use it. But where it's going is that it'll be available to everyone in the world, right? And as it becomes available across the enterprise to non-technical folks who were traditionally in the customer engagement, customer experience world, we think it'll unlock a ton of enterprise value. We think of these people who are crafting intelligent customer engagement using AI as the data artist. I think everyone talks about data scientists in some ways, the people who can create, you know, take that customer lens, create stories that are compelling, but be able to use data and AI, I think that's where the world is going. I love the data artists. You got data engineering, a hot area. Data science will still be there, but you don't need to hire a bunch of them. They're going to be more of the magician behind all the magic and the art. They're going to be almost supplying the paints, if you will. The canvas is the business, and the artists are creating the solution. You said it best. That's amazing, yeah. DJ, thanks for coming on theCUBE and congratulations on the announcement. And again, we'll keep in touch. Love the vision. You guys were ahead of the curve. And again, this is a whole big opportunity. JNAI is really teed up initially right out of the gate for marketing and data. That's what you guys do. Very excited about the opportunity ahead, and thank you for having me here. Okay, this is a CUBE AI conversation here in Palo Alto. I'm John Furrier, your host. Thanks for watching.