Log In to continue learning
Data Engineering
Intermediate
Topics
Data Engineering
Data Processing
Deep Learning
Search and Retrieval
How's it going, Carly? It's going amazingly. How are you? Doing well, doing well. Good to see you. Yeah, you too. Yeah, awesome. So for the learners out there, this is my friend Carly Taylor. For people who don't know who you are, do you want to give a quick introduction? Oh man, I don't think we can do it quickly. In a nutshell, I work on machine learning at Call of Duty, and I'm also a full-time content creator on LinkedIn and Instagram. Awesome. Good content. And one of the things that we're going to talk about is career advice. You, I think, fondly dispense career advice to noobs and to new people and experts alike. But I guess maybe walk through your journey into data. How did you get into this field? Well, I'll start by saying some of my advice is better than others, so take it at your own risk. But I got into data science like a lot of people with STEM backgrounds did. When I was finishing my graduate degree, a lot of my friends were getting hired by FinTech to do this new role called data science to help basically with quantitative analysis and trading at FinTech firms. This was like back in the day when this was really new, and I thought it was intriguing and it sounded really exciting and like a good use of my skills because getting a job as a chemist, come to find out, was kind of difficult. And that was your background is chemistry? Yes. Okay. Interesting. And then you got into data science? I did. Okay. How did you go about doing that? I actually started as a marketing analyst for a marketing company. They were looking for people who had the kind of technical skills that I had, and they were willing to overlook the fact that I knew literally nothing about marketing. Perfect. It was a great combo. But I think this is one of these approaches to getting a career in data that often happens, right? Where you probably aren't a domain expert, but maybe you have some sort of quantitative skill or engineering jobs and then you get hired and so forth. But how did you put some time into learning about marketing in order to improve your odds of doing your job? You know, there was a lot of on-the-job training. It was a marketing consulting company, and the way that they approached things, it was extremely proprietary. And so I think that helped me a lot because they were aware of and used to having to do most of the training for their specific field on the job. A lot of what I did to prepare, I don't think actually ended up helping me. If I had it to do over again, I would have taken some more serious business classes and tried to learn a little bit of the jargon ahead of time. Like, what is a P&L? I had no idea what any of these things were, right? I had a very hard science background, and so it was a lot of sink or swim, Google on the fly. Yeah. And then you got to learn all the business lingo, like circling back. Oh gosh, yeah. And then you're like, oh no, they're circling back. What do I do now? It happens a lot. It happens a lot. So, and now you're, you know, kind of fast forward. Now you're doing machine learning at Microsoft or Activision for Call of Duty. What's a day in the life of an engineer on your team like? You know, it really depends on what kind of engineer that you are. I would say that for my machine learning engineers, their day starts with stand-up, where they go over, you know, what they've been working on with our project managers, and then they'll dig into whatever problem they're looking at. Most of this is going to look like, I'd say, EDA, exploratory data analysis is probably 70% of their time, just to try to get a grasp on what we have and what we don't have, what we need. Then I'd say, you know, 20% is modeling, and then 10% is getting it into production and monitoring. But most of it is just literally getting your hands dirty with data. Yeah, it happens. And walk me through this. Do your engineers work with data engineers? We have some dedicated DevOps folks. So I would probably, at another company, call them data engineers. I'd say that at our company, I think they were called software engineers, but they do mostly DevOps. We have some folks on our team who can flex into that role. Our machine learning engineers are also expected to have some data engineering chops. And then, you know, where that handoff happens in the pipeline is just depending on who feels the sense of ownership any day of the week. It's always a nebulous space. It seems to be. I think the divide, and it's one of the things I want the learners to realize, too, is in the real world, job titles are very, they can be fuzzy in some cases. As you point out, software engineers on your team might be doing data stuff. Machine learning engineers might be doing data stuff as well. Data engineers might be doing machine learning stuff. In your mind, what's the difference between a data engineer and a machine learning engineer? I would say that a data engineer would not expect to be able to build an end-to-end machine learning model. I would expect them to be able to dig into data from a standpoint of knowing that it is, you know, factual from a point of there's nothing happening in the pipeline. There's no droppage of data. There's no transformations that are nonsensical. But I don't expect them to understand the data in terms of, like, real world impact or business impact, right? The machine learning engineers, it's their job to sanity check the data in that way, in a business context, and then to also build predictive models on top of it. So I'd say availability, you know, is more data engineering. And then understanding, to me, is more machine learning engineering. Interesting. And I guess, maybe, can you give the learners some examples of the types of machine learning that you're doing with Call of Duty? Yeah, it really, really depends on the problem. Off the top of my head, you know, we use computer vision. We do a lot of anomaly detection because I work in security. So anything you can think of, you know, that big banks do, like, you know, fraud analysis and detection, looking for anomalies in the data, looking for outliers in data sets, that's kind of our bread and butter, really. Interesting. And you've worked your way up from individual contributor to now overseeing a team. Can you walk the learner through that progression of being an I.C.E. to now being a leader? Yeah, this is such a fun question because it's always really contentious, I think, you know, and it's a question people have of, you know, when should I switch to management or should I? I got into data science because I love the technical aspect of my job and it's scary to kind of relinquish that in terms, you know, to chase more of a management position. And so, for me, the process looked like stepping up when there was a lack of leadership in a certain space and we just needed it. So I'd say that my movement into management came out of necessity, as opposed to a hardcore plan that I had. But you'll often find in your career that when you flex into spaces where there might be, you know, some sort of something is lacking is when you grow. But when you find yourself in a place you might not have thought you'd go into, because you're kind of just acting out of a need that you see. What's one piece of advice you would give your younger self in terms of career growth? Give yourself a little bit more grace to not know things all the time and not feel pressured to constantly be up to date on everything that happens in data science and machine learning. I stressed myself out a lot and burned myself out a lot when I was first starting, trying to keep up with the latest and greatest all the time in every area of data science. And that was exhausting because the field is so broad. So I would have picked a specialty probably sooner. Now that I'm in the security world, you know, the anomaly detection world, that area is my area of specialty and I like to keep up to date in that one specific area. But when I was younger, I didn't understand the idea of like a T-shape level of understanding. You know, I was going for breadth instead of depth. Can you explain the T-shaped concept to the learner? Of course. Yeah. So if you think of the field of data science as being, you know, extremely vast along some sort of x-axis, depending on how wide that is, you can only go so deep on any given topic. Right. If you aim for breadth, you're going to go width wise. If you aim for depth, you're going to go vertically. Right. And I think a good data scientist has a T-shaped level of experience where you're you might know a lot about a lot of different things, but you don't have depth on any one of those topics except for the one area that's your area of expertise. And that's where you choose to go deep. And so your expertise might look more like a T-shape and it'll be centered kind of around that area that you find is your area of expertise. So you might even not know anything about something that's way over here because your T just doesn't extend that far. And that's OK. Not everyone has to know everything about everything. It turns out. It turns out. Yeah. Well, there's a lot to learn, too, especially right now with A.I. It feels like there's new advancements every every day. So keeping up on top of white papers, new products is impossible. It's impossible for people who are trying to get their first job in data. Do you have any advice or tips? Have your resume looked at by at least five different people. And don't have them be the same people every time. If you find that your resume is not working, change it. Don't be so precious about anything on your resume that you feel unable to strike it if it's not working. I see a lot of people, especially people who are switching careers because I've been there, feel unable to let go of accomplishments in their prior career because those were career defining moments for them. Right. I had a lot of amazing publications as a chemist and top level journals. Guess what's not on my resume anymore? And I had to strike because that didn't necessarily matter to the people that I was interviewing with. Now, could I bring that up in an interview and say, also, yeah, I'm published in some top ten journals. Sure. And that's actually really great to bring up. But in terms of resume space, was that telling the story that I needed to tell or was it just me feeling unable to let go of these past accomplishments? Because when you're flexing into a new space, you don't really have those accomplishments anymore. And it's hard to feel like such a newbie. But sometimes you just have to let that go. What if you have no accomplishments? Well, then it's easy. You just yellow your way through. It happens a lot in the job market right now, you know, as we're filming this, it's always competitive, especially for new people, I would say it's acutely competitive these days. How do you advise the people listening to this to stand out in a very competitive job market? I've been thinking a lot about this. And so for job seekers, I will say there's a new trend for submitting videos ahead of getting an interview. There's a lot of platforms now that will ask you to answer questions on video and submit that to the company. I see a lot of people complaining about this and not wanting to do it. And I do understand when you are overwhelmed and you have a lot going on, sitting down and recording a video is the last thing that you want to do. But I will say, use that to your advantage, because that is your opportunity to stand out from the rest of everyone, to not just read off a script, to look polished and have nice lighting. You know, like 99% of these videos that I see are low effort, low quality, that people just don't that you can tell that they don't care. And you can tell that they're reading from a chat GPT answer to the question that was provided to them. If you do even just a little bit more than that, and like put on a nice shirt, you're going to stand out. And so if you start doing, you know, even if it's 5% better than the median of people, you'll start to see yourself pull ahead. And so you don't have to be at 100% all the time. You just have to be better than most people. That's such great advice. Yep. It's like the, you know, the advice of like, how do you survive a bear attack? Like, you don't have to be the fastest person in your group. You just don't want to be the slowest. Right. That's so true. Any other advice you want to give the learners with respect to careers? I think the fact that you're already taking this course shows that you are taking yourself seriously, and you're doing everything you can to set yourself on a path for success. Be kind to yourself. Give yourself grace to not always be performing at 100%. Even on the days that you show up, that's better than 0%. And it's hard to get into something new. It's hard to teach yourself something new. And it's even harder to convince other people that you've done all of this work and that they should take a risk and hire you. And so if at any moment any of that feels overwhelming, it's okay. You'll get through it. Just keep showing up. Awesome. Well, thanks, Carly. It was great talking to you. And I hope the learners learned a lot. And definitely thankful for your time. So thanks. Yeah.
