 Live from Stanford University, it's theCUBE, covering the Women in Data Science Conference 2017. Hi, I'm Lisa Martin, welcome back to theCUBE. We are live at Stanford University at the second annual Women in Data Science Conference. It's a one day tech conference and we are joined by Julie Yu, who is the founder and chief data scientist of Pymetrics. Julie, you are on the customer panel today. So welcome to theCUBE. Thank you, great to hear. You have, it's great to have you, such an interesting background. Thank you. Neuroscience meets engineering or engineering meets neuroscience. We'd love to, for folks to understand a little bit more about those two, how they're combined and also about Pymetrics. But give us a little bit of a background as a woman in the sciences, how you got to where you are now. So as you mentioned, my background's in computer engineering and I went into PhD program in ElectroPoint and computer engineering because I wanted to study artificial intelligence. I was fascinated by the notion of artificial intelligence. So my research topic started in automatic speech recognition system as in building computers to decode and decide for human speech. And after a couple of years, I got sort of frustrated with just the engineering approach or statistical methods based approach to improving the existing speech recognition systems that are out there. Because I thought myself, we're trying to make computers understand human speech and mimic human function when we don't really understand how our brain works. And I don't really know exactly what happens when you listen to you speak, when I listen to you speak and when you listen to I speak. Like what is going on? We didn't really have a good sense. So I wanted to study neuroscience. So I quit engineering and I went into PhD program in neuroscience. And there I started doing a lot of neuroimaging study, just looking at human cognition and just figuring out what is going on when people perceive and process these signals that are out there. And was your idea to eventually marry the two? I didn't really think about it that way, but it just sort of happened. As in like my background in engineering did sort of hold me into doing some of the projects that I did when I was doing my PhD and my postdoc. And while I was doing all of that, I just sort of evolved to be a data scientist without really me realizing I was doing everything that a typical data scientist would do. And this was even before 2008. The job title of data scientist wasn't even around then. So it sort of happened because of where I came from and because what I was interested in. And as I was doing that, like it just ended up being a good marriage. And then there it was. Talk to us about, tell people what Pymetrix is and what the genesis of this company was. Right, so Pymetrix is a platform that uses neuroscience based games and data science to promote predictive and bias free hiring. So how it became a product was because I was going through postdoc and my co-founder was also going through business school and we were both going through the phase of, okay, we don't want to stay in academia. What do we want to do with our lives? And at the time we realized a lot of the career advising tools that are out there were not scientific and they were not data driven. And we felt that there is a clear need for a tool that can actually use all these data that are out there to help people figure out what they should be doing with their lives. So I thought, we thought that we were uniquely positioned to use our background in engineering and neuroscience and build a product that could actually solve this challenging problem. And that's how we started Pymetrix. That's fantastic. You started about three years ago in 2013. So really getting rid of some of the biases, share with us what some of the biases are is it test scores, SATs, MCATs, GPAs? Well, there are many, many different kinds of biases in the hiring process right now. I think there is a preconception of what an engineer should look like and I think that plays a lot. And when you do going to an interview, like how you look and how you dress, that adds to the bias. There's ethnic bias, there's gender bias and there is bias based on test scores and what school you went to. So we want to sort of remove ourselves from that and really get down to what kind of person you are and are you really, I guess, have the right set of skills to succeed in certain job functions. And we do that by measuring yourself. Instead of taking your subjective answers from questionnaires, we do that by objectively measuring your behavior and these games are based on neuroscience research. So we know that they actually measure things that we want them to measure. Like for instance, your ability to pay attention, your risk appetite on all those things that we think matters as to what makes you good at certain things and not so good at some other things. So we use these objective data and data science and predictive modeling to come up with predictions as to how good you will be in certain career versus some other career. Really incredible need for that. So it's game based. So it's an actual game that people will play that will help understand more of who they are as a person, their behaviors, those patterns. Tell us a little bit about the invention of the game. What was it like? Who was it for? Right, so the games were actually sourced from neuroscience research community. We did not create these games. What we did was we actually just took them from research and medical settings and applied it through hiring. And we know that these are relevant to measuring your attributes and your personality. So why not use it for hiring and just career advising? Because it makes sense. We're trying to measure your qualities, your soft skills and whatnot. Why not just use it for something that could really benefit from these sort of data? What we did though is we actually made these games. They're not really called games in research community, but we made it shorter and we made it more applicable to the things that we are trying to use it for. Because you took feedback from some of your early adopters who were saying, maybe it's taking me too long, maybe some of the recruiters might say, they gave you some very viable feedback that have helped you optimize the products. Right, I mean, as a data scientist, I always think the more data, the better. But that also means that people would have to sit in front of their computers and play an hour-long battery of games. A lot of people were thinking that it might be just a tad too long and companies felt that spending 45 minutes to an hour could be a discouraging thing and people felt to take effect and we could see that in the results. So we ended up making it shorter. We went from 20 games to 12 games and we cut it down to 25 minutes long and I think now we're in the sweet spot where we do get enough data, but at the same time we're not making it an hour long. Right, so this is really targeted for people coming out of university programs, whether it's bachelors, masters, doctorate, et cetera. And also what type of companies who are looking to hire, what's kind of your target market for that? So I think mostly Fortune 500 companies because a lot of these companies do hire in large volumes. So it helps to have us go to these companies and build their models based off of their employees. And if a smaller company comes along and they only have 10 employees in the job function, then it's extremely difficult for us to build a model based off of their 10 employees. Whereas if it's a larger corporation, then we can have 200 employees play and we can build a model based on their data. So yeah, so generally large corporations is our target client. I'm curious in terms of some of the data that you were saying that you were analyzing, are you seeing, if we look at data science as a great example of the event that we're at, report from Forbes recently that said it's the best job to apply for in 2017. We're looking at now what's going to be happening predicted over the course of the next year and that's a shortage in talent. Are you seeing with some of the data that you're taking in, are you seeing things that are mapping to that, like people that are really geared towards that? Or are you seeing more companies that are looking for computer industry data science type roles? Is that increasing as well? I think companies are definitely looking for more data scientists and I think also people are figuring out that there are data science programs like graduate school programs and I think the supply of data scientists is definitely increasing but at the same time or more so, the demand for data scientists is increasing and not to mention the available data that's out there is increasing at a faster rate than anything else. So yeah, I mean it is I think the best time to be a data scientist right now. And let me ask you one more question about looking at skills. So we have such a great cross-section at this event of leaders in retail and obviously what you're doing in the neuroscience gaming merging world. We've got professors here. It's such in data science is such an interesting topic. It's obviously very horizontal from a skill set perspective kind of the traditional skills of being a statistician, mathematics, being a hacker. A lot of things that we've been hearing around the show today and really aligns with what you're doing is more on the behavioral insight side of you have to be able to communicate what you're seeing and be able to apply it. I'd love to understand the profile of an ideal data scientist that you guys are seeing from your data. What are some of the other behavioral attributes that maybe you're some of the non-teachable things that you're seeing that really come up that this would be a great career path for someone? I think personally, I think like intellectual curiosity is number one and they would have to have strong self-motivation and discipline because you could love like analyzing data and you could just be doing that for how many days I don't know. And that said, you need to be to actually come up with a good story. You got to be a good storyteller. And if you have artistic flair to make the data vis like beautiful, then even better. But it is important to go from beginning of the project where you have a bunch of data set and actually come up with actionable results that people can use. And you're not only always going to be communicating with the data scientists. So you do need to be able to present your data in a more succinct and easily digestible way. That sounds like, you know, as the chief data scientist for Pymedrix, that's who you're looking for to hire on your team. Give us a little bit, last question here, just a little bit of an overview of what your data science team looks like at Pymedrix as you're helping to leverage this data to give people opportunities with careers. What does your team look like? Our team has very diverse backgrounds. We have a few PhDs in physics and we, well I have PhD in neuroscience and there's another data scientist with PhD in physics. We actually have one guy measured in data science and we have another guy who measured in bioengineering. And yeah, so it's definitely diverse background but general theme is that you do need a good quantitative foundation. So whether it's engineering or physics, it is helpful to have that statistical or analytical mind and if you can actually apply that then I actually love solving problems and I think data scientists is a right role Sorry for interrupting. You're on the career panel at Woods 2017. Is that the advice that you would give to kind of the next generation of kids that are interested in this but aren't quite sure what industry they would want to go into? What industry? I think if they're even remotely interested in going into data science I would encourage them to pursue it. I think it is one of the most fascinating fields right now and there's never going to be a shortage of needs for data scientists. So if you like it, if you think you are going to be pretty good at it I say go for it. Fantastic. And you've got a great audience here. This is being live streamed in 20 cities. I think across the globe are 75 cities. I have to get the stats right. But there's a big opportunity here to be an influencer and we thank you for spending some time with us today. Best of luck on the panel. Thank you. Thank you for watching. We, I'm Lisa Martin. We are live with theCUBE at Women in Data Science 2017 hashtag WIDS 2017. Stick around. We'll be right back.