Top Strategies to Advance your Career in Data Science Panel

according to a survey at the u. s bureau of labor statistics they have projected that data science jobs will grow for more than 20 34 from 2016 to 2026. the same source has described these careers as the fastest occupational growth in the mathematical area of the industry you know what we would all do well to put the focus on these types of careers and that's exactly what we're doing with this panel discussion today so with that said i'd like to invite our panelists to join me on stage and i'll introduce them first of all welcome

virla van leem put an entrepreneur data scientist and full stack developer she's a managing director and head of data science and analytic health a london-based company that's on a mission to accelerate innovation with health care by leveraging health care data birla is a strong advocate of the r programming language and she's passionate about sharing her knowledge when it comes to using r at the enterprise level next we have greg kokiyo greg is an industrial engineer with more than 10 years of experience bringing people processes systems technologies and data together with the goal of optimizing for the highest business returns as a technology manager he owns an ai roadmap that ensures customers are ordering products that are safe while at the same time exceeding their expectations all right then we have uh nikon rasamoff excuse me nikon rosimov rahay he's joining us this morning for our panel discussion um and uh let's see he um sorry um he uh he's a product manager at facebook ai more than 10 years of experience in building b to c and b to b startups from the ground up he holds a phd from cambridge university of all things in computational neuroscience as well as affiliations with mit and singularity university as an expert in information driven product design his publications and patents deal with how to minimize vulnerabilities resulting from sharing too much information we could all use some help with that nikon's product portfolio includes semantics cyber readiness security scorecard automatic vendor detection symantec cyberwriter cloudflare bot management along with other various insurance and security analytic

platforms at this time nikon is responsible for the privacy and developer experience of ai data and feature engineering at facebook so wow with such an incredible um i don't know you might call them distinguished panel here this morning i'm sure we're going to do justice to the subject at hand so let's jump right in all right let's see our first question of the day all right what um in your opinion okay what would be your best career advice for aspiring data scientists okay and um from an educational perspective what would you recommend to a person that's just graduating from college let's start with greg what's your take yeah thank you for uh having me joe uh and i'm honored to be uh among uh such giants today it's a true honor um so data science is in a very um interesting space i think uh the good thing about this one is you don't have to necessarily have uh an academic degree to enter that realm i remember one of my good friends harper zahota said something that really stuck with me he described how sometimes you have other specialties like an accountant would need to have a certificate to prove that they can be professionals out there i think the same thing can apply to a data scientist is that they can take advantage of if they can't go to a university they can take advantage of these online trainings to get certificates but that's not all they have the opportunity to take a look at these projects or come up with some projects to build their portfolio so they can prove that they can learn and apply these skills at a uh a professional environment so uh that's one thing that i can open people's eyes on these inspiring professionals and then the other thing too is i would hope that while in college they were kind of training themselves on how to network if they didn't do it that's okay as they graduate i think one of the focus they need to to uh one of the things they need to focus on is how to network and while they work on building their portfolios networking and learning how to contact people and get their attention it's something that's going to get them inside in the door all right very good uh great answer greg uh i'd like to get uh oscar perez's take on this as well um let me uh introduce oscar perez i we had some technical difficulties but he's

here he's held executive level roles in several leading companies oscar brings decades of experience in international business relations importing and exporting multi-million dollar telecommunications equipment negotiating with foreign governments to achieve business objectives and working efficiently under extreme duress in several countries across south america central america and the caribbean during his tenure at t merrill lynch and other large firms he has directed numerous successful upgrades extensive turnkey projects efficiency improvements major network installations and other revenue generating initiatives whether you're talking about construction performance enhancement network architecture day-to-day operations packet switching fiber optic management or high level analysis and contract management oscars command additional communications and technical savvy in three languages has set a high standard to follow for many years and we hope that language today will be english so oscar uh what's your take on this first question from an educational perspective what would you recommend for a person just graduating from college well uh thank you joe for that introduction i don't know what else to say about that uh regarding that i would mention that go to graduate school and the reason i say that is that there are many programs out there to help the uh the young person to be able to get a handle on what to do in industry and the program that i would say is the most competitive is the mba program i find that of the graduate programs available the mba program is the number one and if you look at the top ten and you looked at the other nine all of them combined don't measure up to what the mba program gives and the reason i say that is i've got an mba myself all of our kids my children my adult children have graduate degrees our daughter for instance her name is monica she currently is director of counseling and medical services at a major university she would not have gotten that job without it uh our son who got his mba from unc university of north carolina in chapel hill he is now the um uh chief commercial officer at doodle so the whole idea behind this is that that's how you get involved in what you need to do on an industry perspective all right excellent thank you oscar uh virla what's your take what's my take um so like if you're just graduate then um it can be quite overwhelming because there are so many things out there you feel like you're just from school and you don't know anything yet what i would recommend is that if you're just graduating that you should focus on learning programs so for example you have a lot of traineeships that you can follow so instead of being thrown out into the world you can follow a traineeship where you actually learn how to how it is to work to begin with but also learn the skill that you might need in the field and if you choose a traineeship and you do it at a big company then you can actually explore that company um while still being on the job and i think learning on the job is one of the greatest things you can actually do because it's well win-win you're working and you're learning um so i would definitely recommend for every graduate to look at these kind of programs because it's a great step up to begin your career indeed it is dr nikon how do you see it that's right um well even though i am a doctor i just want to make sure that everyone understands this are just my personal opinions so i'm not representing anyone here um yeah my perspective on this is actually has been changing recently um normally i always thought you know like you have to work really hard through the things that you're worst at uh but actually i've seen a very successful examples where you just like focus on the things that you're best at and i think that's one of the things that i think distinguishes a lot of the employers instead of like keep pointing you with your nose into the things that you're making mistakes on and saying no fix that like the school right when we go to school a lot of that is about hey you've made a mistake here actually in the professional life you become much more successful if you focus on the things you're good at and passionate about because it creates like a virtual cycle and you spend more time on it it's just my interpretation why it works but it definitely works and so my suggestion is um see what you're good at and start doubling down on it like i give you a couple of examples we have lots of engineers and they have like some sort of a little passion about data science and those engineers are actually the ones that have a much bigger impact because they come in they are passionate about data science they take a couple of courses and then they become really good data centers because they already have strong engineering backgrounds so all of this data science that they do actually goes directly into production so i think that is one of the tricks you don't have to focus on the things you're bad at focus on the good things all right excellent i like that okay next question um

um so the person with the degree as we just discussed uh they've received you might say the hard skills right the basic framework the technical know-how and so forth verla what um what soft skills do you see as being needed to be successful here yeah so often when people talk about data scientists the first thing you think about is fancy machine learning algorithms programming languages all these hard skills yeah but actually when you're a data scientist and especially in the smaller companies you're not only the programmer you're also an analytics translator you are somebody who needs to be able to translate questions that the business asks to you to actually analytical solutions so you need to be able to talk to people to get really through um what is the problem here and what am i trying to solve and those communication skills those soft skills you will need and also you can develop a very shiny machine learning model which is fantastic but if you can sell it to your colleagues like why they should use it or why it is of value then your machine learning model isn't worth anything so you should also be able to present whatever you made um and i think those are actually the most important soft skills you yeah you would need in a data science role um if not any role really in a company you need to be able to communicate and to present um in order for your work to well actually land absolutely uh greg what's your take on this yeah i think i fully agree with varela she uh nailed it and i think you know when it comes to soft skills you have to definitely include communication and if you look at a data science project's life cycle throughout all the stakeholders ask is that we have some clear communication to make sure that uh no drifts is uh is happening one of the things that i advise data science to have is to train themselves on how to be great detectives uh how to really ask the good questions to understand these business problems because that will save them a lot of times now the other thing too is how to be good influencers and to be a good influencer you have to empathize you have to you know uh lead the stakeholders in the direction you would like but for that you have to really understand where they're coming from so influencing with authority is something that's very key uh in my opinion excellent uh how about you oscar wow these guys really nailed it i'll tell you if you're able to articulate or talk the talk articulate what you know what's in here and be able to express it in such a way that not only other people understand it but they see that hey that makes a lot of sense and it's going to work and they're more likely to want to implement the issues that you have brought forth with the solutions that you've also come up with absolutely nikon how do you see it yeah i really like what greg said because i think empathy is the key here right when you have empathy you start understanding the perspective of the people you're working with and so let me give you a brief overview of what i think is important uh when you understand the perspective of data scientists that have been in the industry for a while it has been an uh kind of a development like an evolution initially the whole data science thing was all about give me really cool results i have business impact problems come change this find the solution to this problem everyone was just about business impact and then slowly it evolved to oh wait we actually need to put it in production and so a lot of this was about scaling data science and putting things in production then once you put it in production a lot of the customers started asking wait why did i get rejected so it became all about explainability and like trying to understand what actually happens in the black box then after the explainability has been kind of addressed customers were like hey what's happening with all my data right so privacy became a big issue right and then the next stage of the evolution was wait i'm noticing that my friends get rejected but i don't get rejected the whole fairness thing came out right and so there is an evolution happening and i think the last stage of this evolution is what i call machine learning experience it's about making sure that everyone who's participating in machine learning really can participate and influence and understand what's happening it's not only the data scientists as you kind of pointed out it's the business people the engineers the designers the legal team right and so this whole machine learning experience needs to be so that everyone can participate it's about inclusion that's kind of my thoughts excellent all right well what have we done we have established that they have both hard skills and soft skills right

you might say that they have pretty decent credentials oscar what do you value most experience or credentials and um and why well speaking from experience in my last assignment with atnt i was in cyber security and uh you have to have experience you cannot afford to have someone uh come in and the way i look at it they gotta hit the street running when you have a cyber issue you don't have time to think about possibilities you've only got time to implement some well-known specific issue not issues but uh um things to implement or to mitigate the circumstances that you're facing so experience is uh monumental in my world nikon um what's your perspective yeah i have a pretty controversial perspective um and again it's uh really mine uh i i'm not trying to represent anyone here but the way with the way i see it is at least from credentials perspective there is a little bit of a bias issue because you kind of you recognize people or they went to the same school or they're probably as good as i am right um and so it kind of excludes a lot of people that are probably really good they just happen not to be at the same school um and experiences it's kind of very similar right oh this person was a part of the competition that i was part of or like they did the same course right it all comes down to you start recognizing people uh and identifying yourself during the interview because you're seeing them doing the same things as you did and that obviously worries me a lot and so i actually believe that some of the companies in the past or currently are kind of moving out of this and they're really focusing only on a standardized interview with standardized interview questions and i think the reason why they do it is obviously we understand their downsides but one of the reasons why they do it is because it removes this bias so they don't look at credentials anything that you've actually done and you have you actually have real useful experiences they have standard questions and they ask those questions and of course there are downsides but if i were from a kind of interviewee perspective i would really focus on that focus on the interview that's my answer to that okay all right um let's hear what uh

greg has to say about this yeah i it's uh to me it's a very hard question to answer because it will vary with the uh company who's hiring right um if like oscar mentioned there are some industries that value experience because they don't have time to train or to ramp up that training uh they need somebody to come in and bring you know already established framework to address specific issues when a startup might look at a list of credentials that shows that sends a signal that this person has the skills or many skills to address many problems that may come with managing a startup so it really depends on the environment and i also like what nikon is saying is establishing a standard process where you can remove these biases is probably a good thing you could say somebody coming in with experience will also require more money right so a startup may uh be saying oh well i'm going to value the other so it really depends on the work environment uh the culture of the company in what they really uh are trying to achieve and what they value uh as a business i love it um i think you give the best answer that any technical person will ever give to either a technical question or a non-technical question it depends right that's always my favorite answer all right okay so as a follow-up to this question all right let's suppose that we're in a situation uh in which experience is valued above credentials all right let's flip it on its side you know sometimes that poses a conundrum doesn't it right um a catch-22 if you will how can you get the necessary experience if and when you don't even have a job to begin with let's start with nikon on this one since you have the controversial stuff that's right but i'm going to maintain the controversy yeah go ahead go for it my friend we're both doctors we can do it yeah that's right yeah no i really think that i see a lot of people that have amazing credentials our amazing experiences and they come to the interviews and actually i had the same problem i came in you know i'm a doctor and i'm like hey what is this question why are you asking me this really stupid question what are your three weaknesses right like it's a classical question and then you get angry about it because you're like i'm way above this and much better than this right why are you asking right and i really think this humility that it's an interview and they're gonna ask you very you know questions that you might have opinions on but you really need to nail that interview and all the different industry subjects you can see like for example engineering right engineering has a very standardized interview now called the lead code questions right it's always the same questions about 100 of them and you need to be able to nail them uh product management has the same thing nowadays right it's very standard questions uh you have to develop a product sense metrics and execution right data science has the same thing right and so you really need to acknowledge the fact that you're going here for an interview it doesn't matter how amazing you are that's what i would focus on just nail the interview yep you gotta nail it all right

um what do you think oscar oh my goodness you know one of the best ways to get this uh to get that experience that on the job experience and i still i have to speak from my perspective in my industry is that if you have or if you find the uh a school that provides you with the ojt training or on the job training right that or internship program and that's the reason i mentioned the my solution of getting a graduate degree because there are many schools out there that have some tremendous training programs available that give you the experience by working with other organizations working in the industry while you're getting your degree and that way you did both worlds you're able to have that experience and you're able to get your education and at the same time you might even be able to get a job with the same place in the same place where you are doing your uh internship and that's something that i've seen work on and off in my own industry okay yeah well uh it works with a lot of our industries i think and you know so let me focus on something that may seem a little more technical um virula what are the basics before fancy machine learning algorithms and um what should a data scientist focus on from the start of his or her career yeah so data science and fancy machine learning algorithms go hand in hand but people often forget that there is a basis and that basis is very boring maybe for some people but it is statistics and it's basic data analysis skills like visualization understanding your data because most of the business problems honestly and i'm going to rip you off your dream here maybe yeah they are not solvable with machine learning they can be solved with simple data analysis visualization and understanding your data so i think and besides that on every machine learning algorithm the basis is statistics it's just that so what i would really focus on as an aspiring data scientist is also have those basic skills under control so make sure that you understand how your data can be distributed how you should analyze data how you can make the best visualizations and start from there and if you master that then you can focus on fancy machine learning algorithms um that might give you extra insights on top of that excellent uh greg what can you add to this stunning insight that burlap provided yeah uh that was great yeah absolutely i think you know somebody who's very early in their career whatever position they're in i think they should focus on understanding the language across the departments there may be companies where customer churn or customer retention is expressed in different ways so understanding the business language is very key another thing too is understanding the processes and create some sort of uh data flow of that process understanding what the data is originating to see the end to end of these processes focusing on that understanding how dirty the data might be and how to uh you know gather the best practices to clean that data and save you a little bit of time in the process understanding how to communicate right now that you've learned the business terms you can communicate with that data that you're discovering that you're exploring communicate what you found just like verily was saying how to analyze perform some uh fundamental statistics exploratory uh analysis of that data visualizing this and learn how to communicate these findings to the business those are the things that needs to happen uh first before you jump into the uh with the big guns well you know what it sounds to me in my opinion i don't know no matter which way we go with this there's going to always be challenges you know right you know what that's a good thing and challenges i know they strengthen us don't they they build character they prepare us for future successes and um they help us to deal with and learn from failure i like what robert f kennedy once said only those who dare to fail greatly can ever achieve greatly right so with that in mind let's shift gears again all right what would you say was one of your most challenging experiences in your varied career and what did you learn from it uh let's start off with the controversial guy nikon i'm sorry that's right man i'm sorry no you're not controversial i love what you're saying yeah i love it go ahead i have another controversial thing um actually one of the things that i worked on was called automated vendor detection and so the idea here is that you have a company and you want to know who are they using microsoft office are using a mac what where do they store their money which bank they use and this information is obviously very valuable and so as a data scientist you might just jump into it and you say oh i'm gonna kind of analyze some data but not every solution to the problem is always data science i think that's one of the one you know for him everything looks like a nail and so i think it's good to remind yourself like when you have a profession when you when you're like something specialized then focus on the fact that you're trying to solve a problem not that you have something that you want to use right and in many cases it becomes a different type of problem maybe it's a scraping scanning problem if it's an engineering problem cyber security problem is oscar mentioned right so solve the problems don't try to use your tools yeah exactly one could also say eat your own dog food indeed i will that's another conversation for another time all right oscar what's your take on this hey i like what neutron said and i don't think it's controversial at all nope when you need to be this old you know i didn't grow this yesterday it's been going on for a long time i've been in the business for over 40 years and you have the opportunities to get through challenging experiences and one in particular that comes to mind is a project that i managed in colombia and that's uh not a whole lot not too long ago and some folks might even say in a dallasy far far away for me it was anyway i was managing a turnkey project in colombia uh my company was responsible for upgrading their telecommunications infrastructure uh that is into the 21st century uh this would include their wireline that telephones the uh their data the wireless communications or wireless phones you know these guys cell phones and uh not only that but the outside plant requirements the short version they wanted to uh bring their country of colombia into the digital technology from the analog infrastructure that they had so it was a big deal uh columbia is a country that's made up of 32 departments they can call b they can be trial departments it can be called states or provinces so therefore the project was in various stages of deployment at any one particular time that in itself was uh was a challenging experience so lots of things can happen uh it's almost like a juggling ad you know you've got all kinds of balls going up and down in the air and any one of them can fall at any one time hopefully it's not a big one it's not going to kill you so uh here i am happily working in my office in uh the capital city border that when i get a frantic call from the president of my company now i don't know how many of you have ever gotten a call from a president i had never my president so it must have been important so he tells me that frantically that the governor of san andres had proclaimed a cease and deceased order on all the project work that was going on in his island and uh he little took it seriously he said that um not only were our local contractors sent home so this is serious stuff all of our expats that's our engineers were actually escorted by the police to the airport where they actually saw them get in the plane and leave some of them didn't even have an opportunity to pack before they left uh not only that there was more uh all of the materials that we had staged for the job were confiscated and sent back to customs i had no clue i didn't know that to do that i still don't know they could do that uh and not only that all of the um uh heavy equipment that we had in the island to do all this work um was over taken over by the company in other words by the country in other words it was confiscated by the government so i honestly didn't know what to do with that uh my president parting words with me was oscar fix it just do it and i'm thinking geez uh he wasn't alone by the way he had witnesses my boss the uh vp of operations was with him and a whole bunch of other people i didn't even care about who they were but i knew they were going to be looking after me so how do i do this what i learned from this is how do we solve a crisis that placed a 100 million dollar project on the rocks in jeopardy if san andreas failed because of the uh political clout and the exposure that this particular governor had the project in the rest of the country could drill down the tubes as well so we were dead in the water no turning back i needed to turn a loose-loose situation somehow into a win-win so that was my dilemma and dilemmas are the makings of character right yeah right exactly so you know that's uh you might say that that's like a kobayashi maru you know for those of you who are star trek fans that is the no win situation that everybody would like to um uh like to learn how to get out of and uh unfortunately i've been told we're uh we've got a time constraint here so uh one last question before uh before we go to the um uh the questions from the audience um you know we we're talking about data science the challenges that we have and uh the the different things that you got coming at you at um at many you know from many different angles uh you know what the past has been you know what you have what you know what's in front of you what you have to face um

um what um where do you see the future of data science taking us uh we have time for just one perspective uh nearly virile i can't even pronounce it sorry what um uh how do you see that the future of data science so i think data science has now moved on from like this experimental play around field to a more professional field i think data science is moving way more towards devops and how do you actually get all these nice algorithms into production and how are we going to make actual value of it so it's going way more from okay we're pulling all the tools and all the techniques that we have into our company and we just do it into okay a practical solution so how are you going to monetize the value from data science and how are you going to make your solution actually practical for your company that's where we're heading towards excellent well thank you for your perspective virla um so now we have some questions that have been sent to us from our audience uh first of all uh networking in itself is a skill how can we learn and excel in that skill and how can one reach out to people what should we talk with them when i guess about is i think that's what they meant what should we talk about with them when in search of internships or entry-level jobs um who would like to take this one i can take this one all right greg yeah so uh first of all google okay use google uh search uh the best practices of networking um and if you can't find anything uh test the waters uh if there are any trainings uh i'd like to say that networking you have to network with a purpose do so your homework first whoever you're talking to do some research uh figure out you know what that person is is doing uh what field they're evolving in and then um explain why you're reaching out and what you're trying to get out of this and see if you can offer uh even your help sometimes these folks you're reaching out to they don't need help particularly but uh always show your availability uh so bottom line is google use it and make your research and and go test the waters but test it with purpose don't try the same thing every time thinking that you're gonna get the same results you know a technique works for this one the technique might work for the other one and then forget about the things you can't control sometimes somebody's in a bad mood sometimes somebody's on vacation won't respond to you in the next couple weeks move on test the waters time isn't your advantage i like what you said greg about um you know you try something and it doesn't work move on did you know the definition of insanity is doing the same thing over and over again and expecting different results that is insanity we have no room for that in data science excellent all right next question let's see any suggestions for case studies that oscar talked about being old i'm having a hard any suggestions for case studies and where one can practice some before the interviews all right who'd like to take that one i can address that dr nikon let's hear from the doctor i think my perspective on this is the best way to practice is to just do it just apply to lots of places and then start going to the interviews and then you will slowly get the hang of it and you'll realize what kind of questions they ask and then you will start learning more and maybe reading about the interviews like i think that's the trick grow through a lot of interviews and then take notes right uh ask which questions have been asked which question have you done a good job none of the employers are trying to trick you i think a lot of the employers actually give you the questions ahead of time all right so i know a lot of companies they do practice rounds because of the problem with inclusivity right because they know that a lot of people have a advantage because they come from a place that gave them the practice already maybe the university paid for practice rounds and so there is this imbalance that some people don't get this practice runs and so a lot of the employers actually give you practice rounds ask you the same question and then again in the interview so um do all of this yeah i'd like to add one thing something real quick here particularly to the ones who think that they cannot transfer from one position to the other to data science for example uh say for example you're an accountant and you're like oh data science it's kind of hard to pierce through well if you're looking for a case study you are the case study your experience is my case study right take a look at your realm and then create a case study for that and apply data science to it your goal your role or your your goal is to build the bridge between what you've done as an experience with what the employer is looking for so build that bridge be with those use cases or those business cases from your own experience so you can showcase that you can apply those skills to their business needs can i add something to that john yes 30 seconds uh greg thank you very much that was great what i would add to that is you've got friends uh test out do some role play with your friends go over some questions that you might find uh difficult for you to address and get some feedback because that way you're able to actually see how you perform in an interview by performing in an interview with a someone that in a good environment in a safe environment somebody that knows you someone that can give you some feedback excellent all right well ladies and gentlemen thank you so much for joining us today i should say ladies and gentlemen um so where can we find you all on um on social media or wherever uh nikon yeah i think my linkedin handle is nikon and there is an s in the end nikon's that's it all right greg whether you google it or go straight to linkedin if you type in greg coquillo you will find my linkedin page excellent all right oscar same for me you go to oscar perez and linkedin all right birla i'll let you finish this off yeah same for me i don't know if there are many beautiful nameplates here in the where in the world but i'm sure you will find me that's a good point right there villa von liemput i love that man you guys are awesome all right well folks thank you so much for joining us today i appreciate your time and uh we've absolutely enjoyed um this uh this session and uh folks we uh we may not have been able to get to all the questions but um perhaps we can arrange to have the questions emailed out and uh we'll um we'll see how that goes all right thank you so much thank you