 My name is Dave Vellante and with me are two world-class technologists, visionaries, and entrepreneurs. Benoit Dajaville is the, he co-founded Snowflake and he's now the president of the product division and Florian Duetto is the co-founder and CEO of DataIco gentlemen. Welcome to theCUBE, two first timers, love it. Yeah, great time to be here. Now Florian, you and Benoit, you have a number of customers in common and I've said many times on theCUBE that the first era of cloud was really about infrastructure, making it more agile, taking out costs and the next generation of innovation is really coming from the application of machine intelligence to data with the cloud as really the scale platform. So is that premise relevant to you, do you buy that? And why do you think Snowflake and DataIcoo make a good match for customers? I think that because it's our values that are aligned meaning it's all about actually today allowing complexity for customers or you close the gap or the commoditize the access to data, the access to technology. It's not only about data, data is important but it's also about the impact of data. Or can you make the best out of data as fast as possible, as easily as possible within an organization. And another value is about just the openness of the platform, building the future together. Having a platform that is not just about the platform but also a full ecosystem of partners around it, bringing the level of accessibility and flexibility you need for the 10 years ahead. Yeah, so that's key, but it's not just data, it's turning data into insights. But why you came out of the world of very powerful but highly complex databases. And we all know that you and the Snowflake team get very high marks for really radically simplifying customers' lives. But can you talk specifically about the types of challenges that your customers are using Snowflake to solve? Yeah, so the really the challenge before Snowflake I would say was really to put all the data in one place and run all the computes, all the workloads that you wanted to run against that data. And of course, existing legacy platforms were not able to support that level of concurrency. Many workloads, we talk about machine learning, data science, data engineering, data warehouse, big data workloads, all running in one place didn't make sense at all. And therefore, what customers did is to create silos, silos of data everywhere, with different system having a subset of the data. And of course now you cannot analyze this data in one place. So Snowflake, you really solved that problem by creating a single architecture where you can put all the data in the cloud. So it's a really cloud native. We really thought about how to solve that problem, how to leverage cloud and the elasticity of cloud to really put all data in one place, but at the same time not run all workloads at the same place. So each workload that runs in Snowflake that is dedicated compute resources to run and that makes it very agile, right? Florian talked about data scientists having to run analysis. So they need a lot of compute resources, but only for a few hours. And with Snowflake, they can run this new workload, add this workload to the system, get the compute resources that they need to run this workload. And then when it's over, they can shut down their system. It will automatically shut down. Therefore, they would not pay for the resources that they don't use. So it's a very agile system where you can do these analysis when you need and you have all the power to run all these workloads at the same time. Well, it's profound what you guys built. I mean, to me, I mean, of course, everybody's trying to copy it now. It was like, I remember the notion of bringing compute to the data and the Hadoop days. And I think that, as I say, everybody is sort of following your suit now or trying to. Florian, I gotta say the first data scientist I ever interviewed on theCUBE was the amazing Hillary Mason. Right after she started at Bitly, and she made data science that sounds so compelling, but data science is hard. So same question for you. What do you see as the biggest challenges for customers that they're facing with data science? The biggest challenge from my perspective is that once you solve the issue of the data silo with Snowflake, you don't want to bring another silo, which would be a silo of skills. And essentially there is today a talent gap between the talent available in the market or how it is to actually find, recruit, train data scientists and what needs to be done. And so you need actually to simplify the access to technology such as every organization can make it, whatever the talent, by bridging that gap. And to get there, there is a need of actually breaking up the silos, having a collaborative approach where technologists and business work together and actually all put some of their ends into those data projects together. Yeah, it makes sense. Well, Florian, let's stay with you for a minute if I can. Your observation space is pretty global. And so you have a unique perspective on how companies around the world might be using data and data science. Are you seeing any trends, maybe differences between regions or maybe within different industries? What are you seeing? Yeah, definitely I do see trends that are not geographic that much, but much more in terms of maturity of certain industries and certain sectors, which are that certain industries invested a lot in terms of data, data access, ability to store data in the last few years and now reach a level of maturity where they can invest more and get to the next steps. And it's really rely on the ability of certain leaders, certain organization actually to have built this long-term data strategy a few years ago and now start reaping off the benefits. You know, a decade ago, Florian Halverian, famously said that the sexy job in the next 10 years will be statisticians and then everybody sort of changed that to data scientists and then everybody, all the statisticians became data scientists and they got a raise. But data science requires, you know, more than just statistics acumen. What skills do you see as critical for the next generation of data science? Yeah, it's a great question because I think the first generation of data scientists became data scientists because they could learn some Python quickly and be flexible. And I think that the skills for the next generation of data scientists will definitely be different. It will be first about being able to speak the language of the business, meaning how you translate data inside predictive modeling, all of this into actionable insights or business impact. And it will be about how you collaborate with the rest of the business. It's not just how fast you can build something, how fast you can do a notebook in Python or do predictive models of some sorts. It's about, oh, you actually build this bridge with the business. And obviously those things are important, but we also must be cognizant of the fact that technology will evolve in the future. There will be new tools, new technologies and they will still need to keep this level of flexibility and get to understand quickly what are the next tools they need to use or new languages or whatever to get there. As you look back on 2020, what are you thinking? What are you telling people as we head into next year? Yeah, I think it's very interesting, right? We, this crisis has told us that the world really can change from one day to the next. And this has dramatic and profound aspects. For example, companies all of a sudden saw their revenue line dropping and they had to do less with data. Some of the companies was the reverse, right? All of a sudden, they were online, like in Stackout, for example, and their business completely changed from one day to the other. So this agility of adjusting the resources that you have to the task I need that can change using solution like Snowflake really helps that. And we saw both in our customers, some customers from one day to the next where growing in like big time because they benefited from COVID and their business benefited, but others had the drop and what is nice with cloud, it allows to adjust complete resources to your business needs and really adjust it in house. The other aspect is understanding what is happening, right? You need to analyze, so we saw all our customers basically wanted to understand what is it going to be the impact on my business? How can I adapt? How can I adjust? And for that, they needed to analyze data and of course a lot of data which are not necessarily data about their business but also data from the outside. For example, COVID data, where is the state? What is the impact? Geographic impact on COVID, the time and access to this data is critical. So this is the promise of the data cloud, right? Having one single place where you can put all the data of the world. So our customers all of a sudden started to consume the COVID data from our data marketplace and we had literally thousands of customers looking at this data, analyzing this data and to make good decisions. So this agility and this adapting from one hour to the next is really critical and that goes with data, with cloud, adjusting resources and that doesn't exist on premise. So indeed, I think the lesson learned is we are living in a world which is changing all the time and we have to understand it, we have to adjust and that's why cloud somewhere is great. Excellent, thank you. Nikit, we like to talk about disruption, of course, who doesn't? And also, I mean, you look at AI and the impact that it's beginning to have and kind of pre-COVID, you look at some of the industries that were getting disrupted by every talks about digital transformation and you had on the one end of the spectrum industries like publishing, which are highly disrupted or taxis and you could say, okay, well, that's, you know, bits versus Adam, the old Negra Ponty thing but then the flip side is they look at financial services that hadn't been dramatically disrupted. Certainly healthcare, which is ripe for disruption, defense. So there are a number of industries that really hadn't leaned into digital transformation. If it ain't broke, don't fix it. Not on my watch, there was this complacency and then of course COVID broke everything. So Florian, I wonder if you could comment, you know, what industry or industries do you think are going to be most impacted by data science and what I call machine intelligence or AI in the coming years and decade? Honestly, I think it's all of them or at least most of them because for some industries, the impact is very visible because we're talking about brand new products, drones, flying cars or whatever that are very visible for us. But for others, we are talking about different changes in the way you operate as an organization. Even if financial industry itself doesn't seems to be so impacted when you look at from the consumer side or the outside. In fact, internally, it's probably impacted just because the way you use data, the level of flexibility you need, the kind of cost gain you can get by leveraging the latest technologies is just enormous. And so it will actually transform the industry also. And overall, I think that 2020 is a year where from the perspective of AI and analytics, we understood this idea of maturity and resilience. Maturity, meaning that when you've got a crisis, you actually need data and AI more than before. You need to actually call the people from data in the room to take better decisions and look forward and not backward. And I think that's a very important learning from 2020 that will tell things about 2021. And resilience, it's like, yeah, data analytics today is a function transforming every industry and it's so important that it's something that needs to work. So the infrastructure needs to work, the infrastructure needs to be super resilient, so probably not on-prem or not fully on-prem at some point. And the kind of resilience where you need to be able to plan for literally anything, like no hypothesis in terms of behaviors can be taken for granted. And that's something that is new and which is just signaling that we are just getting to the next step for all data analytics. I wonder, Benoit, if you have anything to add to that. I mean, I often wonder when are machines going to be able to make better diagnoses than doctors? Some people say already. Will the financial services, traditional banks lose control of payment systems? What's going to happen to big retail stores? I mean, maybe bring us home with maybe some of your final thoughts. Yeah, I would say I don't see that as a negative, right? The human being will always be involved very closely, but the machine and the data can really help, see correlation in the data that would be impossible for a human being alone to discover. So I think it's going to be a complement, not a replacement. And everything that has made us faster, doesn't mean that we have less work to do. It means that we can do more, and we have so much to do that I would not be worried about the effect of being more efficient and better at our work. And indeed, I fundamentally think that data processing of images and doing AI on these images and discovering patterns and potentially flagging disease where or near that it was possible is going to have a huge impact in healthcare. And as Florian was saying, every industry is going to be impacted by that technology. So yeah, I'm very optimistic. Great, guys, I wish we had more time. We got to leave it there. But so thanks so much for coming on theCUBE. It was really a pleasure having you.