>>Welcome back to Couchbase connect. My name is Dave Vellante and we're going to dig into a customer case study of sorts with two software engineers from a company called trendy all the largest e-commerce platform in Turkey. And with me are in MRA, tan Rivera, both software engineers at trendy. All welcome. Good to see you guys. Hey, before we get into the story, maybe you can tell us a little bit about trendy all. >>Let me answer that question first. Um, tri-annual uh, today is, um, 10 years old. Uh, actually, uh, it starts with them, um, e-commerce company, uh, Jen, uh, especially, uh, for clothing, uh, today, uh, it's serves several, uh, services, uh, mainly still e-commerce right. Um, we, uh, we do our business mainly on technology and we even have a say in technology, uh, technology is our main concern actually. Um, just like that actually now, >>So thank you for that. I mean, you started, I think, I think the company was founded in 2009, 2010. So you weren't, you were just, you know, kind of which we would consider the, sort of the modern era at the same time. When you look back 10 years, you know, major challenges, major advancements from a technology standpoint. So you at, at the time you had a, uh, a legacy database and you, you had a migraine, maybe you could describe the business conditions that drove you to think about actually making a change. What was the before, and then we can get into the after and what was driving that change? >>Um, maybe I could start it a bit. Well, uh, we have a recommendation domain and try new. It's like when you, when you look at a certain product, like for example, you look at a pencil, it it's commanding you, uh, any razor, uh, if you are going to buy a pink dress, it's going to recommend you a yellow dress. So if you're going to maybe buy pants, it will show you some t-shirts according to it. So, uh, since the recommendations, domain group larger, uh, we, we have struggled, uh, to keep it high scale, and it wasn't a relational DB at first, but that's even as product count increased and, uh, our right frequency increased day by day, uh, and our reef performance was affected very dramatically. Uh, I believe. Yeah. >>So you were using a traditional RDBMS, uh, and then, and, and the issue was you quite, you couldn't make the recommendations fast enough. And, you know, we always say what's real time. Real time is before you lose the customer. So you, you have to make those recommendations in time for the customer to act otherwise, you know, what do you do? Send an email after the fact, Hey, you bought this, nobody's going to pay attention to that. Right? You want to catch them in the moment. Um, and so, so what was it that, that led you to, to Couchbase and w and what was the experience of that? You know, whether it was onboarding, you know, the technology, you know, how difficult was it to get up and running to where you are today? >>Um, we were using ch Couchbase in, uh, in inter-annual, um, for several years, and we had experienced on that. Uh, and, uh, we actually, we need performance as described. So, uh, we convert our data structure to, from relational DB to, um, noise, Carol Levy, um, them actually on our recommendation, uh, platform, the main problem was, uh, invalidation process. You know, um, we are selling things and, um, in seconds they can be sold out and we, we shouldn't be recommend them anymore. And we are, we are keeping track of this by invalidation process and relational DB writing those data to our relationship Libby was, uh, was taking two, two minutes too much time. And, um, by changing this structure to, uh, pathways, we, we, we see that benefits, uh, and it takes so, so, uh, uh, short time, actually, >>I'm so sorry if, if I can just clarify w what was taking a long time, the, the updating the actual records, so that you could actually inform a customer that it was out of stock, or was it the coding that was too complicated? >>Well, it was, it was not because, um, there are millions of products intangible, and, uh, those issues are coming huge, actually. So we are keeping track of time if it's sold out or it's, it can be sellable, uh, when, when a product, uh, detail is seen by the customers, we are recommending some other products too, but those other products must be sellable too. So the main, the main problem was that, and, uh, we are writing them in our relational DB. There is a huge rights law actually. So it was not coding. It was the amount of data actually. >>Okay. And so it was the update intensity, um, within the database and the ability of the database to actually return accurate results quickly. So what was the after, like, uh, can you talk about sort of the, the business impact? What were the, the improvements that you've experienced? >>Yeah. Maybe I can ask her that, uh, like parties said that the main reason we switched is because that, uh, there are so many products coming near in trend, and many of them are being stopped being sold out and the updates to it, it was on a relational, the vendor rights, or too much that you couldn't, uh, dur customers that fast reply because the database was getting effected by the amount of high rights. Because when you think about it, there are millions of products coming, and there are millions of rights, uh, operations on the database. So those affecting the reach performance. So, uh, it, it could occur to you that when you click on a product, you would see maybe as took out product as a recommendation, or maybe a product that is not in the website anymore. So, uh, when we switched to, uh, Couchbase that, uh, we saw that, uh, it's using less resources, which, uh, using less posts, active, alive, and it's also, uh, giving responses faster. >>The main reason, uh, we were using relational DB at first was the invalidation process like five. He said, because it was, we had a consumer that was listening to messages, uh, the innovation messages, and then, uh, and then the writing them into database. But, uh, in the part, uh, it meant that actively writing to database that for every product document that you would need to update the document, but for, uh, for, for, uh, for relational DB, it would be vetoed easier to just make this product, uh, every available, false, or true. So that's why we were sticking with relationship with DB at first. And that's why we made it that first as a relational DB, but as time increased and our product count, and our sellers increased, we realized that, uh, we should find another solution to the invalidation process, and we should, uh, switch because, uh, I mean, it CA it has come to a point at one point that it would just maybe, uh, take a solid, so much time that, uh, we were scaling our consumers at nighttime to just not affect daily users anymore. >>Uh, so that's why that's the main reason we switched. And, uh, after switching, we had in, uh, like I said, the response time and high write throughput, and also one of the reasons is also because that the, uh, the application that was with the use of Couchbase because, uh, since strangled is growing larger than our main data centers. And, uh, like we can see that every day, sometimes we deploy our, uh, apps to yet another cluster. And we, that's why we sometimes need to have backups or different data centers, and Couchbase was providing very good relations, very good solutions to this, which is. Yeah. That's why we switched actually. So we asked >>Couchbase running it's if I understand it, it's running the recommendation engine. And do you still use a traditional RDBMS for the transaction system or is Couchbase doing both? >>Yeah, okay. Uh, we are, uh, actually inter-annual, we are in discovery a team, actually, we call it tribe and in discovery, tribe, uh, relational DB, I think, uh, now, uh, very small, uh, small, uh, teams are using it. Um, it's personally just very low actually. Uh, but, uh, other other tribes, for example, orders, checkout, and maybe, uh, uh, promotions, uh, something like other teams are still using RDBMS, but in discovery team, it's very important to serve customers very fast. We need to show them the products immediately. We need to personalize them. Uh, we sh we should, uh, show them, uh, related products in the meantime, in real time, actually. So in this current Stripe, we are, um, barely using it, uh, RDBMS systems, actually. >>How hard was it to migrate from the RDBMS? Because you hear a lot of stories about how difficult that is to do. You've got to freeze the code, you bringing up new code, you've got to synchronize the functionality. How did you manage that? >>Well, to be honest with you, just ask the data science team to just send the products. Uh, at the same time, we were like, we were keeping the legacy API open that the clients were still coming there. And, um, to be honest, there were lots of legs on that, too. So even if, uh, the, the newer products came a bit later, uh, it shouldn't be seen because it was always coming late. So, uh, we had, we made a new API that is connecting to Couchbase and we wanted the data science team to start feeding it, but we asked the clients to switch it by time. I mean, we were still supporting the old one, but, uh, when we, when we asked the clients to switch to the new API, we just closed the last one. So we didn't really migrate any data to be honest. Like we, we, it was from scratch. And since it's a, it's a recommendation domain, uh, we believe it's better to, uh, add data's from scratch because in our new domains, we are storing them in documents. They are always sending a new list to us. So that's how it gets updated all the time. So since it's not a user related data, it wasn't really like a migration process. >>Is this is part of the secret sauce that you're doing. Schema lists, no schema on, right to Couchbase. And is that correct? And how are you handling it? I'm like, how are you getting that awesome write performance? >>Well, the main reason we believe is that, uh, before, when it was relational DB, like for example, loan product to one product and a second product to first product, third product, first of all, that like you were duplicating the records so that when the product gets removed, uh, from, from a product recommendation, or maybe one of, if a product is getting invisible, for any reason, it should be removed, or maybe it could be a stockout that it means it's not that for every record, you are sending your records for invalidation, but in our new system, it means that this, uh, for this content, there are 24 contents let's say, and like four of them that's finished. It's not there. It's okay. You're just replacing the whole list so that you are not duplicating the records. I mean, this is not like first product first and first, the second, and first to third, and first changes you are replicating this, this change three times, like a delete, uh, product one from three, three product, one from two, and you are deprecating the deletion record, but now we are just replacing the list. So you are doing that all of the operation in 1, 1, 1, uh, Kafka queue message. If I should be able, if I was able to, uh, tell about it. So it's a bit hard to explain it in, uh, in speech, but, uh, we have a nice graphic that's showing how we are doing it now. >>That makes sense. Okay. Thank you for that. And so, as you think about, you're modernizing your application infrastructure, where are you at today? How do you see this modernization effort going forward >>Actually, um, today, uh, we are mainly looking for, um, cross cluster replication. Uh, all our products are, uh, uh, deployed, uh, different clusters and different geographical locations. Uh, we, we always using ch um, we try to always use, um, modern products and, uh, uh, try to avoid, uh, old relational databases, especially for our discovery. Right. And, uh, my mandala is modernizing it, uh, all, uh, engineer's keeping up to date with recent technologies and, uh, our customers are happier. They are not seeing some glitches, some, uh, rates, uh, or while they're using our products. >>Okay. So maybe I could double click on that. So, cause you mentioned the impact of customers and I'm interested in your organizational impact and what it means for you internally, but, but when you talk about cross cluster replication, is that to scale, uh, is that a performance impact? Is that for availability? What's the impact of that effort? That modernization effort? >>Uh, I believe it's, it's all, uh, main reason is availability. I believe. Uh, like we can't know when a cluster can go down, we can't be sure about it, uh, in a, in a system we can, but that we should be up and running all the time. And, uh, there should be some, uh, some backups that, uh, that can switch when a cluster goes down. But also the main reason, uh, well, one of the main reasons is to be able to scale because, uh, the, the clusters that we had wasn't enough, uh, considering our user base. So, uh, let's say you want to even extend your user base, but, uh, like the cluster is being a bottleneck to you because you can't get that much users, but, uh, when you do post cluster that you have backup and you have scalability and it's, uh, considering how new considering if the machines are newer, maybe faster response times. I don't know, uh, maybe, uh, network part would know that better, but, uh, yeah, but all of them, I will leave. >>Great guys. Well, thank you so much for sharing your story, uh, uh, MRA and Fati. Uh, appreciate you guys coming on the cube. >>Thanks a lot. Yeah. Thanks. Thanks. Thank you for, uh, hosting. >>Yeah, it's our pleasure. And thank you for watching. Couchbase connect online on the cube, keep it right there for more great content from the event.

>> from our studios in the heart of Silicon Valley, Palo ALTO, California It is a cute conversation. >> Hello and welcome to this Special Cube conversation here in Palo Alto, California Cubes Studios Jon for your host of the Q. We retreat embodies the founder and CEO of H 20 dot ay, ay, Cuba Lem hot. Start up right in the action of all the machine learning artificial intelligence with the democratization, the role of data in the future, it's all happening with the cloud 2.0, Dev Ops 2.0, great to see you, The test. But the company What's going on, you guys air smoking hot? Congratulations. You got the right formally here with a I explain what's going on. It started about seven >> years ago on Dottie. I was was just a new fad that arrived into Silicon Valley. Today we have thousands of companies in the eye and we're very excited to be partners in making more companies becoming I first. And our region here is to democratize the eye and we've made simple are open source made it easy for people to start adapting data signs and machine learning and different functions inside their large and said the large organizations and apply that for different use cases across financial service is insurance healthcare. >> We leapfrog in 2016 and build our first closer. It's chronic traveler >> C I. We made it on GPS using the latest hardware software innovations Open source. I has funded the rice off automatic machine learning, which >> further reduces the need for >> extraordinary talent to build machine learning. >> No one has time >> today and then we're trying to really bring that automatic mission learning a very significant crunch. Time free, I so people can consuming. I better. >> You know, this is one of the things I love about the current state of the market right now. Entrepreneur Mark, as well as start of some growing companies Go public is that there's a new breed of entrepreneurship going on around large scale, standing up infrastructure, shortening the time it takes to do something like provisioning like the old eyes. I get a phD and we're seeing this in data science. I mean, you don't have to be a python coder. This democratisation is not just a tagline. It's actually the reality is of a business opportunity of whoever can provide the infrastructure and the systems four people to do. It is an opportunity. You guys were doing that. This is a real dynamic. This isn't a new way, a new kind of dynamic in the industry. The three real character >> sticks on ability to adopt. Hey, Iris Oneness Data >> is a team, a team sport, which means that you gotta bring different dimensions within your organization to be able to take advantage of data and the I and, um, you've got to bring in your domain. Scientists work closely with your data. Scientists were closely with your data. Engineers produce applications that can be deployed and then get your design on top of it. That can convince users are our strategist to make those decisions. That delays is showing up, so that takes a multi dimensional workforce to work closely together. So the rial problem, an adoption of the AI today is not just technology, it's also culture. And so we're kind of bringing those aspects together and form of products. One of our products, for example, explainable. Aye, aye. It's helping the data. Scientists tell a story that businesses can understand. Why is the model deciding? I need to take discretion. This'll direction. Why's this moral? Giving this particular nurse a high credit score? Even though she is, she has a very she doesn't have a high school graduation. That kind of figuring out those Democratic democratization goes all the way down there. It's wise, a mortal deciding what's deciding and explaining and breaking that down into English, which which building trust is a huge aspect in a >> well. I want to get to the the talent in the time and the trust equation on the next talk track, but I want to get the hard news out there. You guys are have some news driverless a eyes, your one of your core things. What's the hard Explain the news. What's the big news? >> The big news has Bean, that is, the money ball from business and money Ball, as it has been played out, has been. The experts >> were left out of the >> field and all garden is taking over and there is no participation between experts, the domain scientists and the data scientists and what we're bringing with the new product in travel see eyes, an ability for companies to take away I and become a I companies themselves. The rial air races not between the Googles and the Amazons and Microsoft's and other guy companies, software companies. The relay race is in the word pickles. And how can a company, which is a bank or an insurance giant or a health care company take a I platforms and become, take the data, monetize the data and become a I companies themselves? >> You know, that's a really profound state. I would agree with 100% on that. I think we saw that early on in the big data world round Doop doop kind of died by the wayside. But day Volonte and we keep on team have observed and they actually predicted that the most value was gonna come from practitioners, not the vendors, because they're the ones who have the data. And you mentioned verticals. This is another interesting point. I want to get more explanation from you on Is that APS are driven by data data needs domain specific information. So you can't just say I have data. Therefore, magic happens. It's really at the edge of the domain speak or the domain feature of the application. This is where the data is this kind of supports your idea that the eyes with the company's not that are using it, not the suppliers of the technology. >> Our vision has always being hosted by maker customer service for right to be focused on the customer, and through that we actually made customer one of the product managers inside the company. And the way that the doors that opened from working where it closed with some of our leading customers was that we need to get them to participate and take a eyes, algorithms and platforms that can tune automatically. The algorithms and the right hyper parameter organizations, right features and amend the right data sets that they have. There's a whole data lake around there on their data architecture today, which data sets them and not using in my current problem solving. That's a reasonable problem in looking at that combination of these Berries. Pieces have been automated in travel a, C I. A. And the new version that we're not bringing to market is able to allow them to create their own recipes, bring your own transformers and make that automatic fit for their particular race. Do you think about this as a rebuilt all the components of a race car. They're gonna take it and apply for that particular race to win. >> So that's where driverless comes in its travels in the sense of you don't really need a full operator. It kind of operates on its own. >> In some sense, it's driver less, which is in some there taking the data scientists giving them a power tool that historically before automatic machine learning your valises in the umbrella automatic machine learning they would find tune learning the nuances off the data and the problem, the problem at hand, what they're optimizing for and the right tweaks in the algorithm. So they have to understand how deep the streets are gonna be home, any layers off, off deep learning they need what particular variation and deploying. They should put in a natural language processing what context they need to the long term, short term memory. All these pieces, they have to learn themselves. And they were only a few Grand masters are big data scientist in the world who could come up with the right answer for different problems. >> So you're spreading the love of a I around. So you simplifying that you get the big brains to work on it and democratization. People can then participate in. The machines also can learn both humans and machines between >> our open source and the very maker centric culture we've been able to attract on the world's top data scientists, physicists and compiler engineers to bring in a form factor that businesses can use. And today it one data scientist in a company like Franklin Templeton can operate at the level of 10 or hundreds of them and then bring the best in data science in a form factor that they can plug in and play. >> I was having a cautious We can't Libby, who works with being our platform team. We have all this data with the Cube, and we were just talking. Wait higher data science and a eye specialist and you go out and look around. You get Google and Amazon all these big players, spending between 3 to $4,000,000 per machine learning engineer, and that might be someone under the age of 30. And with no experience or so the talent war is huge. I mean the cost to just hire these guys. We can't hire these people. It's a >> global war. >> There's no there's a talent shortage in China. There's talent shortage in India. There stand shortage in Europe and we have officers in in Europe and in India. The talent shortage in Toronto and Ottawa writes it is. It's a global shortage off physicists and mathematicians and data scientists. So that's where our tools can help. And we see that you see travelers say I as a wave you can drive to New York or you can fly to me >> off. I started my son the other days taking computer science classes in school. I'm like, Well, you know, the machine learning at a eyes kind like dog training. You have dog training. You train that dog to do some tricks that some tricks. Well, if you're a coder, you want to train the machines. This is the machine training. This is data science is what a. I possibilities that machines have to be taught. Something is a base in foot. Machines just aren't self learning on their own. So as you look at the science of a I, this becomes the question on the talent gap. Can the talent get be closed by machines and you got the time you want speed low, latent, see and trust. All these things are hard to do. All three. Balancing all three is extremely difficult. What's your thoughts on those three variables? >> So that's where we brought a I to help the day >> I travel A. C. I's concept that bringing a I to simplify it's an export system to do a I better so you can actually give it to the hands of a new data scientists so you can perform it the power off a Dead ones data centers if you're not disempowering. The data sent that he is a scientist, the park's still foreign data scientist, because he cannot be stopped with the confusion matrix, false positives, false negatives. That's something a data scientists can understand. What you're talking about featured engineering. That's something a data scientists understand. And what travelers say is really doing is helping him may like do that rapidly and automated on the latest hardware. That's what the time is coming into GPS that PTSD pews different form off clouds at cheaper, faster, cheaper and easier. That's the democratization aspect, but it's really targeted. Data Scientist to Prevent Excrement Letter in Science data sciences is a search for truth, but it's a lot of extra minutes to get the truth and law. If you can make the cost of excrement really simple, cheaper on dhe prevent over fitting. That's a common problem in our science. Prevent by us accidental bites that you introduced because the data is last right, trying to kind of prevent the common pitfalls and doing data science leakage. Usually your signal leaks. And how do you prevent those common those pieces? That's kind of weird, revolutionize coming at it. But if you put that in the box, what that really unlocks is imagination. The real hard problems in the world are still the same. >> Aye aye for creative people, for instance. They want infrastructure. They don't wanna have to be an expert. They wanted that value. That's the consumer ization, >> is really the co founder for someone who's highly imaginative and his courage right? And you don't have to look for founders to look for courage and imagination that a lot of intra preneurs in large companies were trying to bring change to that organization. >> You know, we always say that it's intellectual property game's changing from you know I got the protocol. This is locked and patented. Two. You could have a workflow innovation change. One little tweak of a process with data and powerful. Aye, aye, that's the new magic I P equation. It's in the workforce, in the applications, new opportunities. Do you agree with that? >> Absolutely. That the leapfrog from here is businesses will come up with new business processes that we looked at. Business process optimization and globalization can help there. But a I, as you rightfully said earlier, is training computers, not just programming them. Their schooling most of computers that can now with data, think almost at the same level as a go player. Right there was leading Go player. You can think at the same level off an expert in that space. And if that's happening now, I can transform. My business can run 24 by seven at the rate at which I can assembled machines and feed a data data creation becomes making new data becomes the real value that hey, I can >> h 20 today I announcing driverless Aye, aye. Part of their flagship problem product around recipes and democratization. Ay, ay, congratulations. Final point take a minute to explain for the folks just the product, how they buy it. What's it made of? What's the commitment? How did they engage with you >> guys? It's an annual license recruit. License this software license people condone load on our website, get a three week trial, try it on their own retrial. Pretrial recipes are open source, but 100 recipes built by then Masters have been made open source and they could be plugged and tried and taken. Customers, of course, don't have to make their software open source. They can take this, make it theirs. And our region here is to make every company in the eye company. And and that means that they have to embrace it. I learn it. Ticket. Participate some off. The leading conservation companies are giving it back so you can access in the open source. But the real vision here is to build that community off. A practitioners inside large formulations were here or teams air global. And we're here to support that transformation off some of the largest customers. >> So my problem of hiring an aye aye person You could help you solve that right today. Okay, So it was watching. Please get their stuff and come get a job opening here. That's the goal. But that's that's the dream. That is the dream. And we we want to be should one day. I have watched >> you over the last 10 years. You've been an entrepreneur. The fierce passion. We want the eye to be a partner so you can take your message to wider audience and build monetization or on the data you have created. Businesses are the largest after the big data warlords we have on data. Privacy is gonna come eventually. But I think I did. Businesses are the second largest owners of data. They just don't know how to monetize it. Unlock value from it. I will have >> Well, you know, we love day that we want to be data driven. We want to go faster. I love the driverless vision travel. Say I h 20 dot ay, ay here in the Cuban John for it. Breaking news here in Silicon Valley from that start of h 20 dot ay, ay, thanks for watching. Thank you.

(upbeat music) >> Woman Voiceover: From our studios in the heart of Silicon Valley, Palo Alto, California this is a CUBE Conversation. >> Hello and welcome to this special CUBE Conversation here in Palo Alto, California, CUBE Studios, I'm John Furrier, host of theCUBE, here with Sri Ambati. He's the founder and CEO of H20.ai. CUBE Alum, hot start up right in the action of all the machine learning, artificial intelligence, with democratization the role of data in the future, it's all happening with Cloud 2.0, DevOps 2.0, Sri, great to see you. Thanks for coming by. You're a neighbor, you're right down the street from us at our studio here. >> It's exciting to be at theCUBE Com. >> That's KubeCon, that's Kubernetes Con. CUBEcon, coming soon, not to be confused with KubeCon. Great to see you. So tell us about the company, what's going on, you guys are smoking hot, congratulations. You got the right formula here with AI. Explain what's going on. >> It started about seven years ago, and .ai was just a new fad that arrived that arrived in Silicon Valley. And today we have thousands of companies in AI, and we're very excited to be partners in making more companies become AI-first. And our vision here is to democratize AI, and we've made it simple with our open source, made it easy for people to start adapting data science and machine learning in different functions inside their large organizations. And apply that for different use cases across financial services, insurance, health care. We leapfrogged in 2016 and built our first closed source product, Driverless AI, we made it on GPUs using the latest hardware and software innovations. Open source AI has funded the rise of automatic machine learning, Which further reduces the need for extraordinary talent to fill the machine learning. No one has time today, and then we're trying to really bring that automatic machine learning at a very significant crunch time for AI, so people can consume AI better. >> You know, this is one of the things that I love about the current state of the market right now, the entrepreneur market as well as startups and growing companies that are going to go public. Is that there's a new breed of entrepreneurship going on around large scale, standing up infrastructure, shortening the time it takes to do something. Like provisioning. The old AIs, you got to be a PHD. And we're seeing this in data science, you don't have to be a python coder. This democratization is not just a tag line, actually the reality is of a business opportunity. Whoever can provide the infrastructure and the systems for people to do it. It is an opportunity, you guys are doing that. This is a real dynamic. This is a new way, a new kind of dynamic and an industry. >> The three real characteristics on ability to adopt AI, one is data is a team sport. Which means you've got to bring different dimensions within your organization to be able to take advantage of data and AI. And you've got to bring in your domain scientists, work closely with your data scientists, work closely with your data engineers, produce applications that can be deployed, and then get your design on top of it that can convince users or strategists to make those decisions that data is showing up So that takes a multi-dimensional workforce to work closely together. The real problem in adoption of AI today is not just technology, it's also culture. So we're kind of bringing those aspects together in formal products. One of our products, for example, Explainable AI. It's helping the data scientists tell a story that businesses can understand. Why is the model deciding I need to take this test in this direction? Why is this model giving this particular nurse a high credit score even though she doesn't have a high school graduation? That kind of figuring out those democratization goes all the way down. Why is the model deciding what it's deciding, and explaining and breaking that down into English. And building a trust is a huge aspect in AI right now. >> Well I want to get to the talent, and the time, and the trust equation on the next talk, but I want to get the hard news out there. You guys have some news, Driverless AI is one of your core things. Explain the news, what's the big news? >> The big news has been that... AI's a money ball for business, right? And money ball as it has been played out has been the experts were left out of the field, and algorithms taking over. And there is no participation between experts, the domain scientists, and the data scientists. And what we're bringing with the new product in Driverless AI, is an ability for companies to take our AI and become AI companies themselves. The real AI race is not between the Googles and the Amazons and the Microsofts and other AI companies, AI software companies. The real AI race is in the verticals and how can a company which is a bank, or an insurance giant, or a healthcare company take AI platforms and become, take the data and monetize the data and become AI companies themselves. >> Yeah, that's a really profound statement I would agree with 100% on that. I think we saw that early on in the big data world around Hadoop, well Hadoop kind of died by the wayside, but Dave Vellante and the WikiBon team have observed, and they actually predicted, that the most value was going to come from practitioners, not the vendors. 'Cause they're the ones who have the data. And you mentioned verticals, this is another interesting point I want to get more explanation from you on, is that apps are driven by data. Data needs domain-specific information. So you can't just say "I have data, therefore magic happens" it's really at the edge of the domain speak or the domain feature of the application. This is where the data is, so this kind of supports your idea that the AI's about the companies that are using it, not the suppliers of the technology. >> Our vision has always been how we make our customers satisfied. We focus on the customer, and through that we actually make customer one of the product managers inside the company. And the doors that open from working very closely with some of our leading customers is that we need to get them to participate and take AIs, algorithms, and platforms, that can tune automatically the algorithms, and have the right hyper parameter optimizations, the right features. And augment the right data sets that they have. There's a whole data lake around there, around data architecture today. Which data sets am I not using in my current problem I'm solving, that's a reasonable problem I'm looking at. That combination of these various pieces have been automated in Driverless AI. And the new version that we're now bringing to market is able to allow them to create their own recipes, bring their own transformers, and make an automatic fit for their particular race. So if you think about this as we built all the components of a race car, you're going to take it and apply it for that particular race to win. >> John: So that's the word driverless comes in. It's driverless in the sense of you don't really need a full operator, it kind of operates on its own. >> In some sense it's driverless. They're taking the data scientists, giving them a power tool. Historically, before automatic machine learning, driverless is in the umbrella of machine learning, they would fine tune, learning the nuances of the data, and the problem at hand, what they're optimizing for, and the right tweaks in the algorithm. So they have to understand how deep the streets are going to be, how many layers of deep learning they need, what variation of deep learning they should put, and in a natural language crossing, what context they need. Long term shot, memory, all these pieces they have to learn themselves. And there were only a few grand masters or big data scientists in the world who could come up with the right answer for different problems. >> So you're spreading the love of AI around. >> Simplifying that. >> You get the big brains to work on it, and democratization means people can participate and the machines also can learn. Both humans and machines. >> Between our open source and the very maker-centric culture, we've been able to attract some of the world's top data scientists, physicists, and compiler engineers. To bring in a form factor that businesses can use. One data scientist in a company like Franklin Templeton can operate at a level of ten or hundreds of them, and then bring the best in data science in a form factor that they can plug in and play. >> I was having a concert with Kent Libby, who works with me on our platform team. We have all this data with theCUBE, and we were just talking, we need to hire a data scientist and AI specialist. And you go out and look around, you've got Google, Amazon, all these big players spending between 3-4 million per machine learning engineer. And that might be someone under the age of 30 with no experience. So the talent bore is huge. The cost to just hire, we can't hire these people. >> It's a global war. There's talent shortage in China, there's talent shortage in India, there's talent shortage in Europe, and we have offices in Europe and India. There's a talent shortage in Toronto and Ottawa. So it's a global shortage of physicists and mathematicians and data scientists. So that's where our tools can help. And we see Driverless AI as, you can drive to New York or you can fly to New York. >> I was talking to my son the other day, he's taking computer science classes in night school. And it's like, well you know, the machine learning in AI is kind of like dog training. You have dog training, you train the dog to do some tricks, it does some tricks. Well, if you're a coder you want to train the machine. This is the machine training. This is data science, is what AI possibility is there. Machines have to be taught something. There's a base input, machines just aren't self-learning on their own. So as you look at the science of AI, this becomes the question on the talent gap. Can the talent gap be closed by machines? And you got the time, you want speed, low latency, and trust. All these things are hard to do. All three, balancing all three is extremely difficult. What's your thoughts on those three variables? >> So that's why we brought AI to help with AI. Driverless AI is a concept of bringing AI to simplify. It's an expert system to do AI better. So you can actually give to the hands of the new data scientists, so you can perform at the power of an advanced data scientist. We're not disempowering the data scientist, the part's still for a data scientist. When you start with a confusion matrix, false positives, false negatives, that's something a data scientist can understand. When you talk about feature engineering, that's something a data scientist can understand. And what Driverless AI is really doing is helping him do that rapidly, and automated on the latest hardware, that's where the time is coming into. GPUs, FPGAs, TPUs, different form of clouds. Cheaper, right. So faster, cheaper, easier, that's the democratization aspect. But it's really targeted at the data scientist to prevent experimental error. In science, the data science is a search for truth, but it's a lot of experiments to get to truth. If you can make the cost of experiments really simple, cheaper, and prevent over fitting. That's a common problem in our science. Prevent bias, accidental bias that you introduce because the data is biased, right. So trying to prevent the flaws in doing data science. Leakage, usually your signal leaks, and how do you prevent those common pieces. That's where Driverless AI is coming at it. But if you put that in a box, what that really unlocks is imagination. The real hard problems in the world are still the same. >> AI for creative people, for instance. They want infrastructure, they don't want to have to be an expert. They want that value. That's the consumerization. >> AI is really the co founder for someone who's highly imaginative and has courage, right. And you don't have to look for founders to look for courage and imagination. A lot of entrepreneurs in large companies, who are trying to bring change to their organizations. >> Yeah, we always say, the intellectual property game is changing from protocols, locked in, patented, to you could have a workflow innovation. Change one little tweak of a process with data and powerful AI, that's the new magic IP equation. It's in the workflow, it's in the application, it's new opportunities. Do you agree with that? >> Absolutely. The leapfrog from here is businesses will come up with new business processes. So we looked at business process optimization, and globalization's going to help there. But AI, as you rightfully said earlier, is training computers. Not just programming them, you're schooling them. A host of computers that can now, with data, think almost at the same level as a Go player. The world's leading Go player. They can think at the same level of an expert in that space. And if that's happening, now I can transform. My business can run 24 by 7 and the rate at which I can assemble machines and feed it data. Data creation becomes, making new data becomes, the real value that AI can- >> H20.ai announcing Driverless AI, part of their flagship product around recipes and democratizing AI. Congratulations. Final point, take a minute to explain to the folks just the product, how they buy it, what's it made of, what's the commitment, how do they engage with you guys? >> It's an annual license, a software license people can download on our website. Get a three week trial, try it on their own. >> Free trial? >> A free trial, our recipes are open-source. About a hundred recipes, built by grand masters have been made open source. And they can be plugged, and tried. Customers of course don't have to make their software open source. They can take this, make it theirs. And our vision here is to make every company an AI company. And that means that they have to embrace AI, learn it, tweak it, participate, some of the leading conservation companies are giving it back in the open source. But the real vision here is to build that community of AI practitioners inside large organizations. We are here, our teams are global, and we're here to support that transformation of some large customers. >> So my problem of hiring an AI person, you could help me solve that. >> Right today. >> Okay, so anyone who's watching, please get their stuff and come get an opening here. That's the goal. But that is the dream, we want AI in our system. >> I have watched you the last ten years, you've been an entrepreneur with a fierce passion, you want AI to be a partner so you can take your message to wider audience and build monetization around the data you have created. Businesses are the largest, after the big data warlords we have, and data privacy's going to come eventually, but I think businesses are the second largest owners of data they just don't know how to monetize it, unlock value from it, and AI will help. >> Well you know we love data, we want to be data-driven, we want to go faster. Love the driverless vision, Driverless AI, H20.ai. Here in theCUBE I'm John Furrier with breaking news here in Silicon Valley from hot startup H20.ai. Thanks for watching.

>> Hey, welcome back, everybody. Jeff Frick here with theCUBE. It's International Women's Day 2018, there's stuff going on all around the world. We're up at the Accenture event at downtown San Fancisco. 400 people at the Hotel Nikko, lot of great panels, a lot of interesting conversations, a lot of good energy. Really about diversity and inclusion and not just cause it's the right thing to do, but it actually drives better business outcomes. Hm, how about that? So we're really excited to have our next guest, it's Vivienne Ming. She's a founder and chair of Socos Labs, Vivienne, welcome. >> It's a pleasure to be here. >> Yeah, so what is Socos Labs? >> So, Socos Labs is a think tank, it's my fifth company, because apparently, I can't seem to take a hint. And we are using artificial intelligence and neuroscience and economic theory to explore the future of what it means to be human. >> So who do you work with? Who are some of your clients? >> So we partner with enormous and wonderful groups around the world, for example, we're helping the Make A Wish Foundation help kids make better wishes, so we preserve what's meaningful to the child, but try and make it even more resonant with the community and the family that's around them. We've done wonderful work here with Accenture to look at what actually predicts the best career and life outcomes, and use that to actually help their employees. Not for Accenture's sake, but for the 425,000 people get to live better, richer lives. >> Right, right. That's interesting, cause that's really in line with that research that they released today, you know, what are these factors, I think they identified 40 that have a significant impact, and then a sub set of 14 within three buckets, it's very analytical, it's very center, it's great. >> I love numbers. I'm you know, by training, I'm a theoretical neuroscientist, which is a field where we study machine learning to better understand the brain, and we study the brain to come up with better machine learning. And then I started my first company in education, and to me, it's always about, not even just generating a bunch of numbers, but figuring out what actually makes a difference. What can you do? In education, in mental health, in inclusion, or just on the job, that will actually drive someone to a better life outcome. And one of those outcomes is they're more productive. >> Right, right. >> And they're more engaged on the job, more creative. You know, a big driver behind what I do is the incredible research on how many, it's called the Lost Einsteins Research. >> The Lost Einsteins. >> Lost Einsteins. >> So a famous economist, Raj Chetty at Stanford just released a new paper on this, showing that kids from high wealth backgrounds, are 10 times as likely as middle class peers to, for example, have patents or to have that big impact in people's lives. In our research, we find the same thing, but on the scales of orders of magnitude difference. What if every little kid in Oakland, or in Johannesburg, or in a rural village in India, had the same chances I had to invent and contribute. That's the world I want to live in. It's wonderful working with a group like Accenture, the Lego Foundation, the World Bank, that agree that that really matters. >> Right, it's just interesting, the democratization theme comes up over and over and over, and it's really not that complicated of a thing, right? If you give more people access to the data, more people access to the tools, it'd make it easier for them to manipulate the data, you're just going to get more innovation, right? It's not brain surgery. >> You get more people contributing to what we sometimes call the creative class, which you know, right now, probably is about 1.5% of the world population. Maybe 150, 200 million people, it sounds like a big number, but we're pushing eight billion. What would the world be like not if all of them, just imagine instead of 200 million people, it was 400. Or it was a billion people, what would the world be like if a billion people had the chance to really drive the good in our lives. So on my panel, I had the chance to throw out this line that I was quoted as saying once. "Ambitious men have been promising us rocket ships and AI, "and self-driving cars, "but if every little girl had been given the reins "to her own potential, we'd already have them". And we don't talk not just about every little girl, but every little kid. >> Right, right. >> That doesn't have the chance. You know, if even one percent of them had that chance, it would change the world. >> So you must be a happy camper in the world though, rendering today with all the massive compute, cloud delivery and compute and store it to anyone, I mean, all those resources asymptotically approaching zero cost and availability via cloud anywhere in this whole big data revolution, AI and machine learning. >> I love it. I mean, I wouldn't build AI, which that's, I'm a one trick pony in some sense. I do a lot of different work, but there's always machine learning under the hood for my companies. And my philanthropic work. But I think there is something as important as amazing a tool as it is, the connectivity, the automation, the artificial intelligence as a perhaps dominant tool of the future, is still just a tool. >> Jeff: Right. >> These are messy human problems, they will only ever have messy human solutions. But now, me as a scientist can say, "Here's a possible solution". And then me as an entrepreneur, or a philanthropist, can say, "Great. "Now with something like AI, we can actually share that "solution with everybody". >> Right. So give us a little bit of some surprise insights that came out of your panel, for which I was not able to attend, I was out here doing interviews. >> So you know, I would say the theme of our panel was about role modeling. >> So I was the weirdo outlier on the panel, so we had Oakland mayor Libby Schaaf, we had the CFO of the Warriors, Jennifer was great, and we talked about simply being visible, and doing the work that we do in AI, in sports, in politics. That alone changes people's lives, which is a well studied phenomenon. The number one predictor of a kid from an underrepresented population, taking a scholarship, you know, believing they can be successful in politics is someone from their neighborhood went before them and showed them that it was possible. >> And seeing somebody that looks like them in that role. >> And so seeing a CFO of the Warriors, one of the great sports teams in the world today... >> Right. >> Is you know, this little Filipino woman, to put it in the way I think other people would perceive her and realize no, she does the numbers, she drives the company, and it's not despite who she is, it's because she brought something unique to the table that no one else had, plus the smarts. >> Jeff: Right. >> And made a difference to see Libby Schaaf get up there, with a lot of controversy right now, in the bigger political context. >> Jeff: Yes, yes. >> And show that you can make a difference. When people marginalize you, when I went out and raised money for my first company, I had venture capitalists literally pat me on the head and treat me like a little girl, and what I learned very quickly is there are always going to be some one that's going to see the truth in what I can bring. Go find those people, work with them, and then show the rest of the world what's possible. >> Right. It's pretty interesting, Robin Matlock is a CMO at VMware, we do a lot of stuff with VMware, and they put in a women in tech lunch thing a couple years ago, and we were talking, and I was interviewing her, she said, you know, I'd never really took the time to think about it. I was just working my tail off, and doing my thing, and you know, suddenly here I am, I'm CMO of this great company, and then it kind of took her a minute, and somebody kind of said, wait, you need to either take advantage of that opportunity in that platform to help others that maybe weren't quite so driven or are looking for those role models to say, "She looks kind of like me, "maybe I want to be the CMO of a big tech company". >> Well part of what's amazing you know, I get to work in education and work force, and part of what's amazing, whether you're talking about parents or the C Suite, or politicians is... A lot of that role modeling comes just from you being you. Go out, do good work in the world. But for some people, you know, there's an opportunity that doesn't exist for a lot of others. I'm a real outlier. I was not born a woman. I went through gender transition, it was a long time ago, and so for most people like me, being open about who you are means losing your job, it means not being taken seriously in any way, I mean, the change over the last couple of years has been astonishing. >> Jeff: It's been crazy, right? >> But part of my life is being able to be that person. I can take it. You know, my companies have made money, my inventions I've come up with have literally saved lives. >> Right. >> No one cares, in a sense, who I am anymore. That allows me to be visible. It allows me to just be very open about who I am and what I've experienced and been through, and then say to other people, it's not about me, it's not about whether I'm happy. It's about whether I'm serving my purpose. And I believe that I am, and does anything else about me really matter in this world? >> Right. It really seems, it's interesting, kind of sub text of diversity inclusion, not so much about your skin color or things that are easy to classify on your tax form, but it's really more just being your whole you. And no longer being suppressed to fit in a mold, not necessarily that's good or bad, but this is the way we did it, and thank you, we like you, we hired you, here you go, you know? Here's your big stack of rags, here's your desk, and we expect you to wear this to work. But that to me seems like the bigger story here that it's the whole person because there's so much value in the whole versus just concentrating on a slice. >> You know, it's really interesting, again, this is another area where I get to do hard numbers research, and when I do research, I'm talking looking at 122 million people. And building models to explain their career outcomes, and their life outcomes. And what we find here is one, everybody's biased. Everybody. I can't make an unbiased AI. There are no unbiased rats. The problem is when you refuse to acknowledge it. And you refuse to do something about it. And on the other side, to quote a friend of mine, "Everybody is covering for something. "Everybody has something in their life that they feel like "compromises them a little bit". So you know, even if we're talking about you know, the rich white straight guy, everyone's favorite punching bag. And I used to be one of them, so I try and take it easy. It is, the truth is, every one of them is covering for something, also. And if we can say again, it's not about me, which amazingly, actually allows you to be you. It's not about what other people think of me, it's not about whether they always agree with everything I say, or that I agree with what my boss says. It is about whether I'm making a difference in the world. And I've used that as my business strategy for the last 10 years of my life, and even when it seems like the worst strategy ever, you know, saying no to being chief scientist after you know, Fortune 50 company, one after another. Every time, my life got better. And my success grew. And it's not just an anecdote. Again, we see it in the data. So you build companies around principles like that. Who are you? Bring that person to work, and then you own the leadership challenge up, and I'm going to let that person flourish. And I'm going to let them tell me that I'm wrong. They got to prove it to me. But I'm going to let 'em tell it me, and give them the chance. You build a company like that, you know, what's clear to me is over the next 10 years, the defining market for global competition will be talent. Creative talent. And if you can't figure out how to tap the entire global work force, you cannot compete in that space. >> Right. The whole work force, and the whole person within that work force. It's really interesting, Jackie from Intel was on the panel that I got to talk, to see if she talked about you know, four really simple things, you know? Have impact. Undeniable, measurable impact, be visible, have data to back it up, and just of course, be tenacious, which is good career advice all the time, but you know. >> It's always good. >> Now when you know, cause before, a lot of people didn't have that option. Or they didn't feel they had the option to necessarily be purpose driven or be their old self, because then they get thrown out on the street and companies weren't as... Still, not that inclusive, right? >> Vivienne: I get it, believe me. >> You get it. So it is this new opportunity, but they have to because they can't get enough people. They can't get enough talent. It's really about ROI, this is not just to do the right thing. >> If even if you look at it from a selfish standpoint, there is the entire rest of the professional world competing for that traditional pipeline to get into the company. So being different, being you, it's a-- I mean, forgive me for putting it this way, but it's a marketing strategy, right? This is how you stand out from everyone else. One of my companies, we built this giant database of people all over the world, to predict how good people were at their job. And our goal was to take bias out of the hiring process. And when I was a chief scientist of that company, every time I gave a talk in public, 50 people would come up afterwards and say, "What should I do to get a better job?" And what they really meant was, what should I write on my resume, you know, how should I position myself, what's the next hot skill? >> Right. >> And my advice, which I meant genuinely, even though I don't think they always took it as such, was do good work and share it with the world. Not just my personal experience. We see it again and again in these massive data sets. The people that have the exceptional careers are the ones that just went out there and did something because it needed to get done. Maybe they did it inside their last job, maybe they did it personally as a side project, or they did a start up, or philanthropy. Whatever it was they did it, and they did it with passion. And that got noticed. So you know, again, just sort of selfishly, why compete with the other 150 million people looking for that same desirable job when the person that you are, I know it's terrifying, it is terrifying to put yourself out there. But the person you are is what you are better at than everyone else in the world. Be that person. That is your route to the best job you can possibly get. >> By rule, right? You're the best you you can be, but by rule, you're not as good at being somebody else. >> It sounds like a corny line, but the science backs it up. >> That's great. All right Vivienne, I could go on for a very long time, but unfortunately, we're going to have to leave it there. I really enjoyed the conversation. >> It was a lot of fun. >> And thanks for spending a few minutes with us. All right, she's Vivienne, I'm Jeff, you're watching theCUBE from the Accenture Women in Tech event in downtown San Francisco. Thanks for watching. (upbeat electronic music)