 Hey, welcome back everybody. Jeffery here with the Cube. We're at the Chief Data Scientist USA Conference in downtown San Francisco and I'm really excited to have Roman Sharkey on the Chief Data Scientist for the MegaStar Millionaire. Welcome Roman. Thanks for having me here. Absolutely. So tell me what is MegaStar Millionaire? Sure, so we're basically the world's first online talent competition platform. You can participate as a performer, you can participate as a fan. We're currently in open beta testing and we'll be going global worldwide sometime early next year and the winner will receive a million dollars is the price. A million dollars. So this is kind of like the voice or America's Got Talent, but it happens all on your phone. Right. There's no no judges. How does the judging work? Well, I plan to appoint a celebrity judge for our worldwide tournament but currently the winners are determined solely by a compound score, like determine how many people voted on the videos, how many people shared this with us. Okay, so you basically open it up and you say people can submit their talent acts, singing, dancing, whatever. Absolutely. For some period of time and then there's a voting happening constantly. It's just kind of an ongoing thing until you hit some gate and then we have to shut it down or come up with semi-finalists or how does it work? Right. So yeah, right now we're running tournaments one by one broken into rounds. So it's about advancing to the next round. It's pretty much fabulous. Okay. And is it kind of by geography or type of type of act or how do people kind of progress up the tree towards the towards the finals? Right now there's no competition within genres. So it's basically about being the best of the best. Okay. And then and then so that's pretty straightforward. Everybody knows the voice in America's Got Talent, those types of shows, but you're doing it with all in the mobile. So what's the chief data scientists do? What is the components of this platform that you're responsible for and that really kind of make it home? So yeah, my role here is two-fold. I'm responsible for the analytics, basically extracting information from how our users interact with our platform and getting some meaningful insights out of that. And there's also the machine learning part is like, for example, one of the major projects that I'm involved on is, I say, obtaining new performers. So it's a system and algorithm that scrapes through YouTube videos and automatically identifies whether a given video has some sort of talent in it or not. And then we can go straight to the person, contact him and ask him or her if they're willing to join our platform. Oh, that's cool. So you can invite me to join even if I don't know anything about the contest just by looking at my YouTube videos and see if there's a match. Absolutely. And what percentage of the people in the contest were found that way versus people submitting their own video? Unfortunately, this is something that I cannot disclose at the station. Okay, that's fair. But I would say the system is already really accurate and it's accuracy is improving. Okay, well, can you tell me how many people are participating in round numbers right now in the first round? Is it or is it the first round still early rounds? Actually, we're running tournament number two. Okay, right. So I would say we have about, I think, 200 performers, 200 performers, maybe 250. I get I get a check out on that. And yeah, so we're deliberately deliberately keeping this number slow. We're not doing any massive marketing at this stage, because we want to just to test out our platform and make sure it's perfect for the world. Right. So is somebody going to win a million bucks on the beta or you're still kind of working out the platform to you know, on the beta you can you can win $10,000. $10,000. Right. And a million will be for the world left. And then where's the money come from? It's a business model just kind of typical sponsors or right. So we've already been funded funded by several strategic investors. Okay. It's about I would say around 10 million in cumulative investment so far. And we're already listed on the Australian Stock Exchange. Where you are. Yeah. So we're inherently an osc. Very good. Like it last you know, a lot of good technology companies come out of come out of Australia. Right. So that's pretty cool. So what are some of the unique challenges from a data science perspective that that you're addressing that you're tackling that really make a difference in in your success adoption kind of how do you measure success and what are you trying to achieve? I'll say the biggest challenge is that basically we're working on things that no one has done in practice so far. So it's about like a pioneering finding ways to accomplish business tasks through data analysis through data science. I believe there's the hardest hardest one. And but on the again on the data science part, I would say it's pretty easy to actually measure the results because like, again, like on my YouTube script apart, you can see what the accuracy of the output is. And I'll say like, if you've done something positive, your accuracy will grow as as is that. Right. Great. So you're here at the conference. What do you think of the conference impressions of the conference? Any any surprises that you've been here? Right. I would say no surprises is really as expected, as really good. I had a chance to meet a lot of really impressive people down here. Yeah. So okay, I'm really impressed. So should people just people log on and submit their submit their YouTube? Where do they go to submit their YouTube so I can win the $10,000 or the million dollars? Sure. We have our apps on the Apple App Store, you know, on Google Play. So yeah, and I can give you money from that. Any favorites that that there's like picking your own kid, you probably aren't allowed to Right. I'm not allowed to influence the results in any way. Okay, we won't put on this bot. Alright, Roman, well, thanks for taking a few minutes to stop by and I look forward to checking it out. I can't can't wait to see what kind of videos and stuff you have up on there. Absolutely. My pleasure. Alright, Roman Sharkey. I'm Jeff Frick. You're watching the show. Thank you. Thanks for watching. All right. Thank you.