>> Narrator: Live from Las Vegas, it's the Cube, Covering NAB 2017, brought to you by HGST. >> Welcome back to the NAB show. Lisa Martin here with Jeff Frick, we have had an amazing Day One. Wrapping up the end of a really informative day, Jeff. I don't know about you, but the, just the theme of the NAB Conference this year being that the M.E.T. effect is >> Right, right. >> convergence of media, entertainment, technology, and so many different types of technology, was really very exciting, so much innovation going on. So much opportunity. And we've talked to a variety of guests today from those who are involved in film and broadcast and lots of different sectors, to sports broadcasting and really just a very, very exciting... I feel like we're at this tipping point of what's going to happen next. >> Right, right. The themes that we see over and over continue. All about democratization of data, all about using data to make your decisions, even within storytelling you want to use data. And there is data that will correlate to certain types of success and not success. A really interesting conversation around how do you build a movie trailer and what percentage of the trailer has the star in it or not, depending on the star, and on who you're targeting with that particular trailer, the answer to that question is different. So, it's a lot of interest. How a cloud is democratized, all this horsepower that's now available to basically anyone if they can scramble up the budget, they can apply the same kind of massive compute power to rendering and other processes as what was exclusive to just the biggest shops before. So it's just interesting how it continues to be the same themes over and over, and it's impacting this media and entertainment industry in the same ways it's impacting travel and healthcare, transportation, IT, everything else. >> Exactly. We talked about before, the data-driven decisions and as we look at streaming services like Netflix, they've got the advantage of knowing everything, and I think we talked about this in the open this morning, everything about us. One of the things that I learned today was they have that advantage, but one of the things they couldn't do until they started creating their own content was change content. You look at the film industry and filmmakers and writers who have historically, it's been a very qualitative intuition-based process, where now they've got data at their power that they can extract more value from and make data-driven decisions. And we're seeing, to your point, across industries that kind of bringing in artificial intelligence, machine learning, leveraging data science to help make decisions that can help really level the playing field for, like you said, some of the big studios that have the money for real-time cloud rendering or had it a while ago, to now some of the smaller ones that can do that and achieve similar economies of scale that they wouldn't have been able to do on their own. >> Right. The other big trend that we see over and over, Lisa, is this idea that before data wasn't always considered an asset. That might be hard for people to fathom that are kind of recent to this world where of course data's an asset. No, data was a liability. It was expensive. I think in one of your interviews, they didn't keep dailies, because dailies were expensive. They didn't keep this stuff. What's interesting in the context of film, if a particular film becomes really important piece of work, you want to treasure it, you want to keep it. You know, we had Sundance on, talking about archiving all this fantastic material, artwork, cinema, whatever you want to call it. So the fact now that in this industry too, because storage is less expensive, but more importantly, they see the value of the data exceeds the cost of storing it, now they just want more storage, more storage, more storage. 'Cause you don't want to delete anything, and of course, it's all generated digitally today in this industry. >> Right, that's a great point that you brought up, where we were talking with the VP of Marketing at HGST, who was talking with one of the major studios, they filmed this scene that was beautifully shot for I think it was a couple hundred extras in the scene, looked back and thought, you know, we should have filmed that for virtual reality. And because they didn't save the dailies previously, they had to recreate the entire thing. So to your point of looking at the value of data, it's now also, you're right, the economies of storage are going down and there's a lot of technologies, flash, hybrid, that are really enabling it to be readily available. But it's also, this data that's now valuable, is creating new opportunities. It's generating new revenue streams. It's something that companies like a Netflix or even broadcast television can utilize to find different ways of providing relevant content to their viewers. >> Right, right. As you said, things to learn. I learned today that, you know, there are so many versions of a particular media asset that are created, for sensitivities that are around a particular country, obviously now for virtual reality, for all types of different playback mechanisms, so they need to keep everything and create many permutations of everything. So again data makes possible, absolutely. And there's a whole 'nother round coming, right, which is all around the analysis of the frame in the video to get the better metadata. And that's just a whole 'nother rash of improvement that's coming down the line. We heard a number of people today talk about all the metadata and how important the metadata is to capture along the process. But it's going to get even deeper in terms of the analysis of the frame level for these pictures, exposing that out, to other kind of machine learning algorithms, sterch, etc., so that it becomes an even better world for the consumer to find, consume and share that which is of interest to them. >> Absolutely. One of the things that I find interesting is how much content is being created by people that probably don't really realize they're creating the content. Everyone's connected. We talked about we had the independent security evaluator, Ted Harrington, on the program today, who was talking about security, not just in the context of media and entertainment, but the fact that it's a very relevant issue. We know it as an issue in lots of other industries. He was actually saying that it is, the media and entertainment industry is actually pretty good, where security, cyber-security is concerned, securing connected devices, where it seems to me that they could be potentially sharing some best practices with some of the other industries that might still think of security as a nice to have. >> Right, right, no. We saw it with Sony, they got hacked earlier, I guess it's been years now, time flies. So security is very important but obviously the hacking of dvds back in the day, which was a big deal. But now it's all digital and you know the windows to make money on these for the big releases, at the big moment, is relatively short. It's a super competitive business. So, security is definitely a very big issue. It's exciting. The other thing that's kind of interesting is the democratization of the power of all these tools. The thing that scares me a little bit, Lisa, and I see this in a lot of big budget movies, is sometimes I think the tech gets in the way of the storytelling. And I think it's a crutch to lean on cool special effects and cool stuff, and forget about you have to tell a story to make it interesting. And if you don't tell a story, it's not. And we talked on one of the interviews today, about even commercials. And we've seen commercials. You know, Coke hasn't advertised "brown sugar water" for a very, very long time, it's all about the emotion of the Coca-Cola. It's about being part of a community. So to start to use actual data to drive the narratives in the commercials when you're not trying to sell a billion dollar movie, you're trying to sell an entire factory production run of a new automobile, the stakes go even higher, your touch points are even lower. So again this whole theme over and over, data driven decisions based on AI, based on measuring the right things, based on knowing your consumer better, because you have to, or else they'll just whoosh, swipe to some other piece of content. >> Exactly, exactly. Yeah I think those were the very pervasive themes that we saw here. But I think there's just tremendous opportunity. It's almost like we're at the tipping point. We had Kevin Bailey on, as well, from Atomic >> Jeff: Atomic Fiction. >> And conductor, and he was saying six years ago, when he had this hunch on cloud where to try to do rendering in real time for big movies like Dead Pool, for example, The Walk, one of my favorite movies, would take a tremendous amount of time. And he said to be able to do this with the speed that we need and the agility and flexibility, a fixed solution is not optimal. So he was really kind of leading edge in that space. And now we're seeing technology as pervasive. But you're right, there can be an overuse of it. So it's really about finding this balance. I think we had a great spectrum of guests on the show today that really showed us all of the different facets, and we've probably just scratched the surface, right? >> Oh, definitely. >> That you can look through to really understand what makes good content, emotional, what makes it successful, and what enables the audience to be in that control of this data that is democratized all over the place. >> Yeah, to get emotionally involved. There's some great lines. It's all about emotion and connecting in a hyper-competitive world for attention. It's really an attention competition these days. >> Lisa: That's a good point. >> It's much harder than it's ever been. >> It is. >> All right, well we've got two more days. >> Lisa: We do. >> So get a good night's sleep. I'll get a good night's sleep. You should get a good night's sleep. We'll be back for Day Two at NAB 2017 with Lisa Martin, I'm Jeff Frick, checking out with The Cube. We'll see you tomorrow. Thanks for watching. (computerized music)

>> Narrator: Live from Las Vegas. It's the CUBE. Covering NAB 2017, brought to you by HGST. >> Welcome back to the CUBE live in Las Vegas at the NAB show. We're having a great day so far. Very excited to introduce you to my next guest, Kevin Baillie, cofounder and VFX supervisor at Atomic Fiction and the CEO of Conductor Technologies. Never a boring day for you with those two titles, I can imagine. >> No, I like to joke that I like to make sure that I always have the most exciting job in the world so I had to pick three to make sure that I never have a down moment spoil that, that day >> Wow, I am impressed. So you just spoke at the virtual NAB conference last month on the visual effects in the cloud, power, and control. Something that I found very interesting was that six years ago, you were kind of on an island going "I have this hunch about cloud." Tell us about, what was that hunch, why did you have it, and what has it generated so far? >> Yeah, yeah, that's a great question. The hunch was less of like, "Hey cloud looks like a great opportunity." It was more of like knowing what wasn't working in the industry as it was at that time. There were all kinds of companies that were kind of like having financial troubles or having a hard time delivering projects, tons of bankruptcies and just really sad stories everywhere. And we looked at the market and said, "There's a ton of work here, this doesn't make sense." Some of the best entertainment is being made right now and it all relies on visual effects, what's wrong? And the further we broke down the problem, the more we realized that like fixed infrastructure within a market that naturally ebbs and flows, it just didn't, there wasn't a match there. So, through that problem, we looked for solutions and cloud was a very obvious one at that point. So we just made the jump. >> And tell us about Atomic Fiction versus Conductor Technologies. Chicken, egg, which one came first? And how are they collaborating together? >> Atomic Fiction came first. It was almost seven years ago at this point that we started Atomic. And we looked for any kind of a way to use cloud. We started using an AWS directly, we then used a tool called Zync. And as we grew, we found that the needs of the company were changing so radically that nothing that was out there could actually keep up with our pace of growth. We had all this customized pipeline that we couldn't find a way to like get it into the cloud. So we built our own and that was called Conductor. And after, I think we were working on like Game of Thrones and The Walk and had just started on Deadpool that we realized it was working so well that we decided to spin it off as it's own company and make a go for actually turning it into a product that could help everybody in the same way that the cloud had helped Atomic Fiction. >> Fantastic, one of my favorite movies is The Walk. I was looking at your website and you think as the viewer, "How did they film this?" You know, this day and age, so much is CGI. Talk to us about what realtime cloud rendering is. How does it enable a movie like The Walk or Deadpool to have that awe inspiring, jaw dropping reaction from the audience? >> Well I think a large portion of bringing that jaw dropping reaction to the audience and that level of realism is being able to run productions in the way that they want to be run. And what I mean by that is, let's take a movie like The Walk where you have to recreate 1974 New York and the Twin Towers, and all these different lighting scenarios. That means we have to build every building, every rain gutter, every hotdog stand in the street down to exacting detail, and that just takes a lot of time. So we spent a ton of time, probably the first three quarters of the schedule just building the city, building the city. And we couldn't render anything at that point And it wasn't only until the very end of the show that we were able to say, "alright, now we have New York is there, let's just put it on the screen." But that takes millions of hours of computing to get that done. The Walk for example, it used 9.1 million processor hours of rendering. That's over a thousand years on a single processor to get it done. So if we hadn't had the cloud, we would have had to been like, "Oh what can we render first "so we don't bottleneck at the end of the schedule?" And really kind of like trying to bend production into the box that we, of fixed infrastructure that we have. But with the cloud, we don't have to do that. We can say, we can go as big as we want to at the very end of the show and get it done if that's what makes sense for the show. Because that's what makes sense for the show, the creative just ends up being that much better. The same was true for Deadpool, the same is true for Star Trek. These movies, they just sort of, you want to craft love into the beginning part of it so the stuff you generate at the end is as beautiful as it can be. >> So is cloud really freeing production from being able to operate in the way that it needs to operate? >> Yeah, yeah, exactly. Because the traditional model is, a visual effects company builds a data center and stuffs it full of computers. In best case, with like three weeks lead time you can like rent a bunch of racks of computers and like shove them in a closet somewhere and get your project done. It ends up being expensive and painful. You need a big team to man all that stuff. Whereas with cloud, we can say, "Hey, I need a thousand computers three minutes from now." And boom, a thousand computers spin up out of nowhere. And the great thing that we've done with Conductor as well is we've gone and negotiated per minute software licensing with Autodesk and the Foundry and IsoTropic and Chaos Group. All these big software vendors in the industry. So not only can you get compute by the minute, you can also get all the software that you need by the minute, right. So you can have three thousand nodes running Autodesk to Arnold, and you, but you run it for 42 minutes and you only pay for 42 minutes of three thousand licenses of Arnold, right. So it's really transformative from a flexibility standpoint. >> And the cost model really flips it on it's head. >> And by the way, the artists get the result back faster. Because you can scale up so big and get the result back to them so quickly without any cost penalty, they see the fruits of their labor while the ideas are still fresh in their head, which is like a huge, like, intangible benefit which has real economic benefits. >> Absolutely, one of the things and themes that we've heard of today is that speed is key. Absolutely critical to whatever is going to happen or whether or not on a shoot, a vision changes direction. And without having the power of the cloud to facilitate something on a dime, there's delays, which all adds up to economic impact. >> Yeah, and you know, back on one of our earliest projects rendered in the cloud, Flight. The Robert Zemeckis movie with Denzel Washington. That exact thing happened, where it was like at the very end, he, Zemeckis realized that he needed this extra set of like a hundred visual effects shots. And if it hadn't have been for the cloud, we would have had to say, "No, sorry we can't do these." "We have to find somebody else to do them." But because the ability of the cloud to accommodate that last minute creative epiphany, we were able to actually do the work. So it really is truly transformative and allowed us to bring in, you know, hundreds of thousands of dollars of extra revenue that we wouldn't have been able to do otherwise. >> Absolutely. In terms of some of the public cloud providers, tell us who you're working with on that end. >> Yeah, so we're working with Google right now, using Google Compute Engine on the back end. And we're actually moving forward with Microsoft and Azure. Adding it as an option later in the year. So, hopefully at the end of the year, we'll be able to support all the large cloud providers. And be able to say, "Hey, Studio X. "We know you have an affinity for Google right now, "but on the next project maybe you need "a very specific GPU type." Or there's a company in China that needs to do some work and Google isn't there. Now Azure is your thing, right. So, I think that the world of cloud providers competing against one another is going to be really beneficial for everyone in our industry for sure. And we want to be there to facilitate a little bit of like, choose whoever's best, right. >> Right, giving you the ability to really be like agnostic on the back end. >> Yeah that's exactly right. >> So as we look at these massive resources that studios are generating, creating such interactive films, what are some of the precautions that you see and you can help them mitigate against leveraging the power of cloud. >> Well, one of the benefits of cloud is you only have to pay for what you use, just like electricity, right. One of the downsides of cloud is you have to pay for what you use, right. So, if you're not careful about the render you put in the cloud or the simulation you put in the cloud, or how long you keep data in the cloud, things can get really expensive really quickly. So, one of the things we did, and this is actually why we kind of spun Conductor off as it's own company. And we just raised our Series A round of funding back in December to build the team out, because a lot of this stuff is really complicated, is one of the big efforts, in kind of a post funding world for Conductor, is on analytics and being able to use data to help people drive production better. So you know, in the very beginning, we have cost limits where you can say, "On this shot, I don't want to spend "more than a thousand dollars." Or, "I never want this artist to be able to spend "more than fifteen hundred bucks a day." But in the future, I think that there is kind of like cloud buzz-wordy things that actually come into real play here where we can use machine learning to detect when things are taking too long and alert people. We can tell people how much a render is going to cost before they even submit it maybe. We can use computer vision to check for bad things happening in the middle of a render before a human ever has a chance to lay eyes on it. So there's all kinds of stuff we can do with data to help mitigate some of the downsides of cloud and hopefully only leave people with like great insights to help them run production better. >> That's fantastic. One of the things that really interests me is the machine learning and the artificial intelligence. To be able to look at whether it's a broadcast outlet or a film studio, to be able to take a look at and evaluate the value and additional revenue streams that can come. But also, in your case, maybe even leveraging AI and machine learning to make certain processes faster thereby lowering costs. >> Yeah, we can actually make proactive suggestions based on, like, you know, thousands or millions of data points and say like, "Hey if you tweak this value on your shading rate here, "you're going to end up with a great visual "and not spend any more time, or actually spend less." So things like that and then also working together with production management systems. Like the guys at Autodesk have a product called Shotgun that deals with schedules and artist assignments. And they can have all the schedule information. We have all the sort of infrastructure information. If we correlate those two data sets together, then we'll be able to actually proactively tell somebody when we think a shot is running behind schedule. Or a shot needs more optimization. And I mean, there's all kinds of things that we can use just purely using data and a trained machine learning model to actually help people run their entire business better, not just an individual shot. >> Right, well, six years ago, when you had this hunch, you said there were some skeptics around there. One, you must feel pretty validated by now, but are you kind of one of the go-to guys, go-to companies of this is how to do it properly? These are all of the advantages, economic advantages, etc, that we can provide? >> Yeah, I think that there were definitely people that told me I was absolutely crazy when I first got started. Some of them are actually using Conductor now, so that's kind of like good. >> That must feel good right? >> Yeah, it's a good validation point and they had a lot of reasons for thinking that we were insane, cause we kind of were. But we just sort of believed deep down that it was going to work. So, yeah, I mean now, I think we're in a great position to help people. And for me, and you know, this is always like a thing that I sometimes get a hard time for, but I'm so passionate about this industry moving into the cloud that I'm just as happy to talk to somebody about how to do it maybe on their own if they're trying to do it on a small scale. Or what our competitors might be doing. Really, through that, I've kind of, we've found a space where we don't really have any competitors yet and we're breaking new ground. Really servicing the sort of medium and enterprise scale customers, and that kind of flexibility and scale and security that they kind of need. So it's sort of interesting in this, in a way, this sort of like selfless, just being excited about cloud has helped us to find a market that we can really and truly add insane value to. >> Wow, that is fascinating. Well, your passion for it is evident. Thank you so much Kevin for joining us on the CUBE. >> Yeah, thank you so much. >> Have a great time at the rest of the show and we'll see you on the CUBE sometimes soon. >> I always do, thank you again. >> Excellent, we want to thank you for watching. Again, we are live at NAB Las Vegas. Stick around. We will be right back.