Kat Gordiienko began her career working full-time as an accountant while completing her senior year of college as an economics major.
Short on time between her job and education Kat was forced to adopt programming on the job. She taught herself SQL and—after a number of subsequent years in BI consulting–ultimately took a Senior Analytics Engineer role with Netflix.
This interview covers Kat’s background and advice she has for accidental engineers like she once was herself:
Max: Welcome, all! Max of “The Accidental Engineer” here. Today, we’re joined by Kat Gordiienko. Kat is coming up on four years at Netflix.
Kat is a really great guest to have on the Accidental Engineer for some really great reasons, including the fact that Kat has a super accidental background of doing what many of our audience members have done which is doing an undergraduate degree not in computer science yet finding herself in a heavily technical career role where the undergraduate degree did not really provide much applied education!
Kat, do you mind introducing who you are, what you do now and how you got there?
Kat: Definitely! Hi, I’m Kat and thank you for having me.
My official title is Senior Analytics Engineer at Netflix, and as you mentioned I’m coming up on four years there.
Prior to Netflix, I worked in BI consulting for about three years, and before that my occupation was in finance and accounting.
So it’s been a very non-traditional, nonlinear path from accounting/finance background to engineering and data. Everything big data is what I do today.
Max: It’s awesome to hear that abut starting out in finance–how did you find your way into programming and working with data?
Kat: I would say a lot of things just fell in place honestly. Even in my first job that I had while I was still getting my degree–in my first real job in a corporate environment, it was in finance and accounting, and I took it because I needed to pay the bills, really, and to pay for school to be able to graduate.
So in my senior year as I was getting close to getting a diploma, I worked for a private equity company in their finance and accounting department, and my degree was in economics. It was pretty close to being similar fields and just working with numbers, that’s what I wanted to do at that time.
We had a very small IT department in that company that basically was a gift and a curse at the same time, and a lot of reporting did not work or make sense especially for end-of- month close. Our financial models would not be sustainable in Excel that we were using at that time.
So we started learning SQL and database basics on our own. In that way, I learned how to code in SQL, how to ask questions from a database and get the answers back.
Max: So “private equity” is a super loaded phrase from a prestige perspective, from a glamour perspective–was working in private equity out of college a super glamorous profession or what made you decide to move on from finance to BI Consulting?
Kat: I really loved the fact that you can formulate a question and then get an answer, and it was all about how you formulate a question. There are so many specifics to that just like in real life that it really pulled me into more of a technical role. And in terms of private equity, I never really thought much about where I would want to end up after college, it was more like plug and play and learn as you go.
And to be honest, everything that I’ve learned today, it’s been on the job. So coming from economics background, I did not take any computer science courses. I’m a heavy believer in learning and a lot of things come to you as you practice.
Max: This is a super common refrain that comes up on our podcast: the idea of getting that first job and then getting paid to learn the skills on the job, which is a really sweet two-way street between you and your employer.
Besides getting paid to learn this stuff, one of the things we often harp on is that many people without a job aren’t sure of what is worth learning.
Max: It sounds like they gave you some pretty strong guidance about what is worth learning in your job at the private equity company.
Did they leave it up to you that SQL ultimately was the best decision for your guys’ problem, or did you have mentors there that pointed you in the right directions?
Kat: I think it’s actually always up to you, even in the job that you’re maybe not happy with, it’s up to you to decide. You can switch teams in the same company or look elsewhere, but ultimately, it is up to you.
And it was a little bit of both, we had mentors and my manager. It was his idea to start exploring SQL and I was super interested in it as well.
And, yes, the idea of getting a job to be paid and be valuable to the company and at the same time, learn something, it’s a back and forth process. I would say nobody ever knows everything and sometimes, especially interviewing for the first job, if you don’t know something, it’s not a problem. Even if it’s your third job or whatnot.
I usually try to illustrate my train of thought, how not knowing everything about a technology or a tool that I was asked about, how I would approach it. And, I think, that’s what all the employers find valuable, illustrating how you would get from A to Z without knowing much about it.
Max: What were some of the most difficult roadblocks you encountered when you were first trying to learn SQL, for example?
Kat: Everything is so public and accessible these days that every time I would have a question, Google is your best friend. To be honest, even today.
If you’re working with a specific tool or specific technology, there are groups and forums of people that collaborate so don’t be afraid to ask questions.
I think my biggest roadblock was me not asking enough questions. I usually want to figure something out on my own and sometimes it takes much longer than if you were to get help, and it’s something that I had to practice a little more.
Max: Do you have any recommendations about how to determine whether a team or an employer is a good team or employer from a learning perspective?
Kat: I usually look at Glassdoor for example, and I read reviews–I read about people’s interview experiences, and people who work for the company or former employees. In my experience it’s been pretty accurate so far, and Glassdoor is not the only resource out there, but kind of like if you’re buying a product you can read reviews on Amazon or whatnot. Same thing with looking for a job. That’s what I do.
Max: Fair enough!
So you first started learning SQL seven plus years ago, are you still using it today?
Kat: Absolutely, everyday.
Max: So outside of SQL, I realize in your toolkit of tools that you guys use on your team at Netflix, you have a tremendous amount and it’s constantly changing.
Do you mind sharing a few of the ones you’re most optimistic about getting to use in the stack in the coming years?
Kat: Definitely. I would say SQL generally is a very non-tool specific language, and anybody using databases and working with databases will, at some point, use it. No matter whether it is Oracle or Hive or Spark which is a new one, Spark SQL, in particular, it’s all the same in terms of writing SQL and the syntax. That has been my bread and butter.
With any data products, there are front-end and back-end tools, right? From the front-end, it’s whatever BI tools are not just hot today but widely used by people, and popular, and accessible, and user-friendly.
Those are the ones that are around but they change very frequently. And at Netflix, we use everything that’s available out there and we have a very wide freedom in terms of what tool to pick depending on the use case, and I love that about my job.
Max: I think many people should be familiar with how Netflix is by and large built on top of Amazon Web Services. Is working with the cloud and Amazon Web Services a significant part of your role? Or are the layers of engineering stacked so that you don’t have to interface with that part of Netflix’s data pipeline?
Kat: There’s definitely some heavy engineering that needs to be done in the upstream jobs and I do not deal with that. But anything that relates to as soon as the data touches the database, I can take it from there.
Max: I know you’ve given talks previously about how Tableau is used at Netflix and on your guys’ team, do you mind sharing for our audience a little bit of a rehash about how you guys use it? What were the alternatives you guys might have used or might still use to Tableau?
Kat: Definitely. When I started about four years ago, we were largely using MicroStrategy and it’s a great tool because it can generate SQL automatically depending on how you set up your schema. We’re still using MicroStrategy a lot, however, Tableau emerged during those four years as a quicker-to-market tool and it’s much quicker for us to build a prototype, for example, for a report. I would say it’s very user-friendly and, to me, it was more like Excel on steroids in a way. And it’s very intuitive to learn.
Even for our stakeholders, we try to empower our stakeholders to just do it themselves and for us to focus on more self-serving tools, for example.
Max: I think your comments about how you work with stakeholders is super interesting. This is a recurring topic on the podcast. Do you mind sharing for our audience a little bit about how interacting with stakeholders is different, coming from BI consulting to being an in-house analytics team member?
What was the interface like with stakeholders and how is that different?
Kat: At Netflix, specifically, we try to empower our stakeholders to do and be able to answer any questions that they want themselves first instead of coming to us for everything. As I mentioned, those who are comfortable with Excel, they can pick up Tableau very easily. What is not easy in Tableau is more customized and specialized things.
In my experience, everything turns into a hack when it starts being intermediate to advanced, and actually my talk on the use of Tableau at Netflix is mostly hacks I’ve implemented personally in the reports that we provide to our stakeholders.
But working with stakeholders in general–in my role I not only collect requirements but I basically build everything from scratch, so I go through all the stages of development, for example. It starts with a business problem, then you decide whether it’s a one-time occurrence or it’s something that will be recurring. If so, then we are going to build a report, for example, that requires probably a lot of ETL work–extract, transform and load–and that’s where I work with the database a lot.
And then, if and when it turns into a report, we’ll go through cycles of improving it or not. And to be honest, in my world reports don’t really live a very long time because things change so frequently. So if a report lives for a year, that’s great.
Max: That sounds like a tremendous amount of time in the grand scheme of things! We’ve actually had one Netflix guest on previously, Jose Moreno. He’s also on the engineering team, although he doesn’t work on the marketing side of things.
For people who are curious about what kinds of marketing problems Netflix has, what are the distinguishing engineering problems that marketing engineering at Netflix deals with in contrast to the rest of the marketing analytics industry?
Kat: I would say we are definitely trying to solve big data issues, and when the data grows exponentially, how do you process it, how to make that processing faster, what are the most efficient ways to aggregate data, what are the most important questions we want to answer with that data, and that is how we would aggregate it and report on it.
What are the missing pieces, because with marketing, a lot of information comes from third-party vendors–how do we integrate third-party vendors in-house and how can we make insights out of first and third party data.
Max: I think, all the database-specific technologies you guys are using is super relevant and that you guys are at the forefront of the industry in a lot of ways.
Are there any questions you wish I’d asked or topics you would like to talk about?
Kat: It depends on whether you want to talk about data processing and BI, right, or general career advice for people who want to get into that role.
Max: I think it would be great to talk about career advice as we definitely haven’t spent as much time talking about it.
For people who might be in finance or accounting at this point in their careers and are interested in how to get to the same place where you’re at in an analytics engineering role: what’s your very broad general advice for those people?
Kat: I would say not being afraid to apply to jobs.
For example, if you read a job description and you don’t match 100% of what the requirements say, you should still apply and I would say if you feel like you only meet maybe 50% of the requirements, that’s a great start.
Even if you get rejected a lot all you need is really one yes to be able to get that job, right?
Definitely, self-education, there are so many materials out there, books and podcasts and online tools and classes that you can take. So doing that, I remember, interviewing for my consulting job that I had, the interviewer, who ended up being my manager, gave me homework to read Ralph Kimball’s book on BI and data warehousing.
Max: Oh man, that’s an old book, right?
Kat: It is an old book.
Max: From the ’80s or something.
Kat: Yes. But it gave me a great…I mean, I didn’t understand half of what I was reading but it gave me a good basis for understanding what is it that I’m about to embark on.
Max: Oh for sure.
Kat: And before starting that job, I actually finished that book. Very theoretical but I have a good understanding of why things work the way they do.
And also, not being afraid to fail because everybody fails and instead, asking yourself, “What can I learn from this?” It’s actually the failures that teach us and get us to the next level. And not being afraid to try. A lot of people are afraid to change things in a data warehouse or in the reports because they’re afraid to break things.
Another tip that I’ve always used is to look how other people did something in the past and try to replicate their code and build on it. That always works.
Max: It’s super funny you mentioned that we…one of our previous guests, Rachel Maltiel Swenson, who leads a analytics team at eBay, mentioned the exact same thing when it comes to working on a problem you haven’t worked on before at your employer. Try to find, for example, SQL queries that co-workers have written, they can talk you through what it does, what all the tables are, what schema, decisions they made, and it’s a hell of a lot easier to jump off from that than it is to go out and try and build up your own familiarity without the guidance of your co-workers.
Max: And one thing, I think, a lot of people get fixated on and I agree with literally everything you said so far, is how people get fixated on paying for an education to be able to teach them what they’re not even sure is worth learning.
For example, bootcamps are a very popular option these days. Speaking to somebody who’s debating that perhaps, what alternatives would you give them to a 12-week, $20,000 bootcamp when it comes to self-educating like you’re advocating?
Kat: Definitely. I think it really depends on if you’re working full time and considering taking a bootcamp, are you required to take time off from your job, right?
If so, I would probably just recommend going and using the tools out there that are online. And honestly, installing the programs or any tools that you need for that job and just trying and building things on your own first.
For me, whatever classes I took, they provide great theoretical knowledge. But oftentimes, the problems that you’re solving there are in a perfect environment. They provide you a perfect scheme of a perfect database where everything is connected and everything exist, which is not the case in the real world.
Working on specific examples that you’re trying to solve works for me better. Even when educating our stakeholders, I usually try to take a problem from our real life experience as opposed to coming up with something very theoretical. So I would do that.
Max: Are there any…I mean, we’ve talked a lot about SQL, we’ve talked a little bit about Tableau, what are some software that you’d recommend people tinker around with on their laptops that’s pretty easy, maybe even free, to play around with to get their hands on SQL or data visualization tools? Is there a free tier of Tableau?
Kat: Gosh, there’s so many tools that you can get for free or at least free trial period.
MySQL, I think, is the easiest to install and create your version of a database. I mean I even do it in my personal life–I have a little database with my expenses, and I have some reports that I put together for my finances.
Max: For sure. Do you use MySQL?
Kat: I do.
Max: Wow, nice!
Kat: I use Tableau…it’s a mix of Tableau and Excel and whatnot, but it’s fun for me to play around and learn.
There’s also a lot of Microsoft MSSQL related tools out there. With Tableau, there’s a free trial version that you can use, and not just Tableau, there are so many other tools emerging these days.
I think I already mentioned that it’s not really about a tool–it’s more about the skills that you’re learning while using the tool and making sure that those skills are transferable because tools will change but the skills and your knowledge is your base forever.
Max: For sure. I think we should take a moment. I realize we don’t have specific roles on your team that we can plug today, but for people who are interested in careers at Netflix and engineering or analytics, do you mind mentioning how people can find this?
Kat: Sure. Netflix is constantly hiring and my department is Science and Analytics. We are always looking for analytics engineers or data scientists. We’re also looking for data engineers, as well. Checking it out on the Netflix jobs site is probably the first resource that I would use.
Max: We’ll include a link in the show notes. One thing I also want to plug real quick is that if you guys have any questions for Kat, check out her profile page on the website. You can submit any questions or comments and I’ll forward them on to Kat. And sign up for our email list, subscribe to the YouTube channel! Kat, it’s been freaking awesome having you.
Kat: Thank you so much, Max.