⭐ Texas educators: our K-5 Tech Apps curriculum is a state-approved instructional material. Learn More.
The STEM Career lesson that matches this interview, “Data Analyst”, can be found in Codelicious High School Computer Science Python. Python is the language that most data analysts use day-to-day, so download a free Python fundamentals lesson to use with this interview! In It’s All in the Details, students will spend time predicting, running, and investigating code to enhance their understanding of output and print() functions.
Name: Theesh Mohan
Title: Product Reliability Specialist
STEM Career Lesson: Data Analyst
Course: High School Computer Science Python
Gather around the campfire (or board room), and learn how storytelling and data analysis converge. Explore a career as a Data Analyst with Theesh Mohan of Meta. Theesh walks us through the process of analyzing data, finding trends, and condensing 10 weeks of work into a 1 hour meeting.
Learn more about careers at Meta: https://www.metacareers.com/
Katie: Successful data analysts aren’t just numbers people; they’re great storytellers. It takes lots of training and skill to look at data, analyze the facts, and communicate solutions to business problems.
Today, you’ll hear my conversation with a pro at the storytelling component of data analysis. Theesh Mohan is a Product Reliability Specialist with Meta, formerly Facebook. He also worked a data analytics companies such as PwC, a big four accounting firm.
Welcome to My STEM Career, inspiring the next generation of leaders. This show is brought to you by Codelicious Computer Science Curriculum; I’m Katie Baird.
In this first section of the episode, we’re diving into questions from our Data Analyst STEM Career Lesson from High School Computer Science, built for grades 9 to 12. Then, we’ll transition and learn more about Theesh’s life, career, and advice.
Katie: Hi, Theesh! Thank you so much for coming on My STEM Career today.
Theesh: Hey, Katie, Thanks for having me.
Katie: We’re going to start out with some questions related to our STEM Career lesson called Data Analyst. And this lesson is from our High School Computer Science Python course built for grades in 9 to 12. So for anyone listening right now, whether you’re watching the video or listening to the podcast, you can find more information about that course in our show notes today. But Theesh, I’d love for you to start us out by introducing yourself. What’s your name, what’s your job title, and where do you work?
Theesh: Sure so, hey everyone, i’m Theesh Mohan and my current job title is a Product Reliability Specialist, and I work at Meta.
Katie: So at a high level, can you explain to us what a data analyst is?
Theesh: Sure, so the simple, obvious answer is a data analyst is anyone who analyzes data. Anyone who works with data and analytics, but perhaps a better answer would be: A data analyst is someone who analyzes data that’s often complex and/or just really large sizes, large volumes, with the intent of answering one or more questions that are typically business-related. Or a data analyst is anyone who analyzes data looking for insights about their customers or their products, or and so on. Or in some cases data analysts is just simply those were looking for evidence or data-based evidence to back up their claims or their hypotheses. So data analyst is anyone who really works with data that’s increasingly complex these days. To answer some key questions or or uncover some insights.
Katie: So you’re saying these data sets are huge. They’re very complicated. Can you walk us through some of the steps that you take as a big data analyst when you’re working on a project?
Theesh: Absolutely. So, the first thing I would say is anyone should do on any sort of project, and then sadly, that a lot of people forget to do is first identify what question or questions you’re trying to answer. Identify what exactly the the purpose of all of this data analysis is and what you want to get out of it at the end. And then after you’ve done that, then identify what kind of data you’ll need. Where do you, where would you get that data from, and how much data do you need? And then what kind of analysis you would do once you have that data.
So figure all of that out first, and once you have that nailed, or at least somewhat sort of you know, confirmed, then you would go out and either acquire that data, or if it already exists somewhere, you would just locate it you would find it, and and, you know, get it into your environment and then clean the data, because, unfortunately, no data that I’ve ever worked with has been of perfect quality. So you first need to clean the data, make sure it’s a ready for the analysis you want to do on it.
And then, finally, you actually do the analysis, gather your results and answer those questions that you set out to answer. And then the final step, often the most important step, is finding a way to summarize those results and present those findings to the end audience, wherever that may be, in a way that’s easy for them to understand and take action on because all the work you’ve done means nothing if someone can’t actually do anything or take action based on it.
Katie: For sure. So what are some of the qualifications that you need to be a data analyst and do that work?
Theesh: So I’d say there’s really nothing in particular, there are no data analyst, you know, qualifications as far as i’m aware, or at least you don’t need anything like that. In general just an affinity, for data and from numbers. definitely helps, and any sort of qualification that gives you a decent understanding of math, specifically statistics. I think that would definitely help, but I can’t think of any must have qualifications that you absolutely need to be a data analyst, because honestly, anyone can be a data analyst if they understand how to look at data, how to work with data, and, most importantly, how to generate answers or insights from data. Math and statistics help, but beyond that there’s nothing particular that I would mention.
Katie: Yeah, yeah, and so that kind of leads into our last question here, for part one, which is: what skills or traits are necessary to be a successful data analyst?
Theesh: There are a few skills i’d say are pretty critical and would help greatly.
So first, I would say, is the ability to understand the high-level question or problem you’re trying to solve and figure out what it is that you’re trying to answer. And then what data will you need to to find those answers. You need to sort of be able to draw a parallel between the question you’re asking and the data that’s out there, or the data that you can get to know exactly how you connect the data with the problem of the question an understanding of some typical data analysis techniques that goes a long way; it’s a pretty important skill to have.
And honestly, you’re never really done learning as a data analyst. This is always a new tool or a new method or something new out there that you you need to learn. And as data sets get larger and larger and more and more complex for time, it becomes vital to sort of keep yourself updated, to use more and more advanced tools and techniques to process or analyze that data. So the ability to learn over time, often just independently using resources on the Internet, that’s a pretty important skill
I would say, and then finally, attention to detail. When it comes to cleaning the data or later analyzing it, often data gets very complicated, and and the answers you present to your audience or the the claims you make are only as valid as the assumptions you make about the data. So paying close attention to the data what it can and cannot give you what you can and cannot draw from it, is pretty important.
And then, finally, like I said, the most important step is at the end of the day being able to summarize whatever you’ve done for the decision makers or leaders or your boss so, being able to clearly and succinctly summarize your your findings, and communicate it sort of to a vast audience, often you know, to those who don’t really understand data that’s gonna be vital.
Katie: Yeah, for sure, some of those skills like communication and teamwork alongside some of those hard skills you were talking about some of the tools and techniques, which we’ll dig more into in part 2. So excited for that conversation. But that was part one, which were questions related to our STEM Career Lesson, Data Analyst, for grades 9 to 11. And now we’re going to transition, and kind of dig more into your particular career journey, Theesh. So i’m really excited to hear about some of these things.
Katie: Those questions drew from our Data Analyst STEM Career Lesson, part of Codelicious High School Computer Science for grades 9 to 12. You can find more information about the course in the show notes. Now, on to the second part of our show. Join me as Theesh talks about his career journey, including real business problems he’s solved using data.
Katie: Let’s just take it all the way back to your education. You got your college degree in aerospace engineering which is literally rocket science. This is the classic joke, right? This is what you have your degree in. And then you moved on to a role in data analysis. So what were some of those skills that you learned in your degree that helped you make that jump between career paths.
Theesh: It’s a good question, and it really was a jump. It was a very different world between what I studied and what I ended up doing for my first job out of college, and I think some of the things that really prepared me for you know, a job in data and data analysis is first the ability to understand a problem at a high level, even if it seems vague at first. The ability to understand the problem at a high level, but then break it down into smaller, more sensible parts, more easily understandable parts, and then dig deeper and deeper, until I was able to, until I’m able to identify the root cause. No matter how my vast or complicated or strange a problem may seem when you first look at it, if you follow an organized structure to approach to breaking down that problem into little pieces, and until you get to the root cause that goes a long way, and essentially that organized, structured approach to problem solving is something that is kind of bread and butter for an engineer. So it’s something that was emphasized throughout my education as an aerospace engineer, and honestly, any kind of engineering would have to thought you that so that was very directly transferable skill, and it helped me understand complex data related problems and and analyze data to solve those problems.
That’s one. The other is I think throughout my I guess college education, I needed to know when to ask for help, and more importantly, where to look for help, because you can’t always expect help to be given to you in the form of a lecture slide, or a textbook, or things like that. So, knowing how to just go off and look for help on my own and just learn on my own was was a pretty vital skill that I I think, helped me a lot as I began my career, because once you’re out in the real world, you know, once you’re deep in the weeds of you know a 20 GB data set, there’s often, you know, there isn’t a guidebook for you to open and flip through. So you need to just go off and do your own research and bumble around to the Internet and spend a lot of time going down paths that don’t help at all until you eventually find what you’re looking for.
And eventually, you often find what you’re looking for only by trial and error. You just try out a few things and you think you’re getting somewhere, and you don’t but then resetting starting again from scratch, and then and repeating that process until you get to your final answer I think that it’s something that I got very comfortable doing through my education, and it was a very directly transferable skill.
Katie: Yeah, it sounds like perseverance is so important in any engineering degree which really then translates so well to your work in data analysis.
So once you got into this field, tell us about some data analytics projects that you’ve worked on – what were the problems that you were trying to solve, and what were some of the ways that you went about analyzing that data to solve the problems.
Theesh: Sure, the one example I often like to talk about is something I worked on, perhaps a maybe a couple of years ago or so. We were essentially we were trying to solve – we were trying to solve a data quality issue, or we were trying to solve some data quality problems for a client that had a restaurant business. So this was a large restaurant chain, with restaurants across the country. and the world, indeed, and what we were trying to do was figure out some issues they were having with their sales. Things like, How often do we have duplicate orders that you know someone tried to place an order on the Internet, it went through. they didn’t realize it went through, so then they called in the same order, so, you know, the kitchen ends up making 2 orders, and you got to throw away one. We lose money on 1. So they were trying to identify cases like this where the the data was not flowing through as it should, and as a result they were losing money, quite literally, wasting time, losing money, wasting food, and so on.
This was just one example. but there were, there were a long list of questions that we were trying to answer, and cases that we were trying to identify, and to do that we had to dig through several several months, sometimes years worth, of restaurant order data which, as you can imagine, for you know a company that had over a 1,000 branches or a 1,000 outlets, it gets very complicated very quickly, and you’re working with very large volumes of data, and often that data is not even of good quality and different branches, send their data in different formats, so it got pretty complicated, so that in in that case I think I had to along with my team, use our full skill set of, you know, data analysis
I guess techniques and methods and skills to really first, bring all that data together, harmonize the data, make sure that all the data fit well together, so that you can look at the entire company as is one large data set. Then, we had to to break down the problem into little little pieces, and then we have to use a number of you know, at the time for me, it was advanced and unknown tools and techniques to to look through that data, to wrangle the data, if you will.
So we did all of that. At the end of the day, We did all that work, and we had to summarize the results on a dashboard. So we had to basically design and build a dashboard where we summarized I think it was 8 or 10 weeks worth of work into a one-hour meeting. And you know, to do that, we had to think very hard about every single, every single square inch of that of that dashboard, and make sure that it made sense to someone who didn’t have the benefit of working with the data for 10 weeks.
So that was really interesting where we’re looking through order data, sales data, customer data, branch data, things like that to identify cases where there was a real financial impact that the company was actually losing thousands of thousands of dollars because of data issues or process issues like duplicate orders or incorrectly allocated gift cards.
Things like that. So that was a really interesting project, and it was, it seemed like a long time, but 10 weeks is actually not a lot of time at all when you when you’re talking about going from nothing to a final dashboard. So that was a really challenging, but but also really interesting project, and we’ve had a few others like this for different kinds of clients, automotive clients, and let’s see, power energy utilities clients, airlines things, like that. So everywhere you go. Every problem we approach was obviously completely unique, and and every industry has its own nuances, its own oddities, if you will.
But at the end of the day data is data, no matter where it comes from and what industry comes from, data is data. And principles of data analysis and the techniques you use, and that the process you follow is going to be more or less the same, no matter where you go.
Katie: Yeah that’s such an interesting real world application, because all of us have been on sort of the customer side of that, you know, ordering food and turning on the lights, but actually thinking about these everyday things that we just take for granted from the perspective of a business is just such a cool part about being a data analyst.
Theesh: Exactly, yep.
Katie: So we’ve talked a little bit at a high level about the techniques that you use when you’re organizing data. But what is the specific business analytics software that you use when you’re looking at a big data set?
Theesh: Good question, and the answer, as is so often the case depends on what kind of data you’re looking at.
More often, it depends on how much data you’re looking at and I would say, Excel, good old Excel, is still one of the most reliable, widely used versatile tools out there to work with data and analyze data in the past, you know, whatever 20, 30 years of Excel’s been around, I can’t remember how long. But Excel is actually been adding features at such a steady pace that it’s now capable of doing some really really cool stuff really advanced stuff to work with data and and process and understand data. So Excel is something that I think we’ve all used at some point and we all will continue to use for a long time to come. So, knowing how to get beyond just some basic spreadsheet functionality to really take advantage of all the features Excel has to offer, I think, is very, very important, and will take you a long way.
But then, once you start getting into larger, and larger data sets even excel, can’t keep up, and so then some of the tools we typically use, one you’ve probably heard of and you will always hear of is SQL or S-Q-L, Structured Query Language. That is something that has been a mainstay of any sort of data analysis effort. for as long as but as long as I can remember and SQL analytics are incredibly powerful, it’s very simple to understand, and it’s it’s built to be easily usable by anyone. And if you take the time and effort to understand SQL and and learn its various, I guess, nuances, because there are different versions of it, you, you can do a lot with SQL, and that is something I’ve used as well quite a lot in my in my job in the past.
And and then, finally, there are times when you want to do some really cool analysis. Things like data science or machine learning things, like that, or even just more advanced data analysis that you can’t do with SQL or with Excel And then Python is what I’ve used. And Python for data analytics, as you’ve probably heard already, since your classes on that, is extremely capable of doing a lot of really cool data analysis really complex data analysis now, and it’s because it’s entirely up to you what you do with Python. There really is no limit. You can get very creative with Python. You could build more and more complex solutions, on Python that still perform really well, even when you’re dealing with immensely large data sets.
So those 3, I would say, are 3 of the more common ones that at least I’ve used. I know there are plenty out there that I haven’t had the chance to work with yet. But Excel, SQL, and Python if you have those 3 in your toolbox. we there is probably not much you won’t be able to do.
Katie: That’s really cool. So just kind of zooming out a little bit, what is your favorite part of your job?
Theesh: I’d say that’s basically coming up with answers to really tricky or or complicated business problems using data in really creative ways. So I really enjoy being given a problem that at first makes absolutely no sense, and that that seems so vast and so insurmountable. and then being able to break that down into smaller pieces and then look at data in really interesting ways and innovative ways, creative ways, and then using those along with some assumptions of course, but but using those innovative approaches to looking at data, to to answer complex problems, I think that’s I think that really intellectually engaging and then really fulfilling and enjoyable.
Katie: And then on the flip side, what are some of the biggest challenges that you face in data analytics for business?
Theesh: Without a doubt, the biggest challenge that we always always always face when it comes to looking at data is first finding the data we need and in a clean, usable condition. I don’t think I’ve ever seen a data set anywhere in my job is of great quality that I can just pick up,
Look at and just analyze from from, day, one it’s always a challenge to first find enough data and find the right types of data that you need, and then find it all, or rather, clean it all, to be in a usable condition.
So, before you can start doing any analysis, or finding any answers, you need to make sure that you have all the data you need in a usable condition, and that’s almost always a challenge.
Another one is like I said at the end of the day what you’re trying to do is communicate whatever you find to an audience to to a decision maker or decision makers, and often it’s hard to find a way to present all the data that you have in a way that’s easy for people to understand. But that still has a required level of detail so it’s it’s. it’s often challenging to find that that sweet spot between getting too detailed and being able to back up every claim you’re making but also summarize it enough so that you can explain it within a half hour or an hour.
So, finding that the right sweet spot can can be a bit of a challenge.
Katie: I love that the challenge that you mentioned, also just ties into what you love the most about your job. That it’s tough to communicate the answers, but also it’s the most rewarding and fulfilling when you’re able to find and share those answers with different stakeholders in a business.
Theesh: Yeah, the harder the problem is, I think, the more satisfaction you get from solving it.
Katie: Definitely. So our last question here today is, what advice would you give to students who are interested in following your career path?
Theesh: Sure. So I think the the most important thing to do to get good at data analysis and to get good at managing data is just spend time playing around with data. There really is no other shortcut to it. Spend time with data sets that, you know, there are hundreds and thousands of data sets out there that are just publicly available that you can use for for practice.
But just spend time playing around with data to get comfortable with data exploration. Just look through the data, try and summarize it in different ways. Try and filter it in different ways. Try and come up with width percentages of populations that certain things apply to or for find – Another really fun way to do.
Some data analysis is, Look at historical data for sports like baseball or basketball, and find interesting answers to, you know, what people have achieved looking at, not just the last year or 2, but 30, 40, 50 years of data and just playing around with data and getting comfortable with exploring data, sometimes even without an end result or an end goal in mind is really helpful that over time you’ll find that that that that makes you more creative and helps you find really cool ways of looking at data and coming up with insightful answers to really difficult questions and often it’s when you’re playing around with data without really any ambition or goal that you start you start picking up on interesting ways you can look at data and start picking up on, maybe counterintuitive ways of filtering data or interesting assumptions you can make to to what you’re looking at that you can then transfer those – you can transfer that learning to to actual problems. When you, when you look at you know data sets with with a goal in mind.
So just be curious, you know. Spend the time put in the hours to just play around with data and explore. Be curious, and be willing to just continuously learn and learn independently.
So like, I said often or almost always the answers are out there somewhere, and you don’t even know where they are, no one’s gonna come give them to you, You just need to, you know. Go out and just look around. Try and find things on your own. Try things out, fail, try again. Fail again. Try again, until it finally works. So, be be willing to just learn independently and learn continuously that that learning never really ends.
And the last couple of things, I’d say one pay attention to detail. And don’t be afraid to to really get down into the very weeds of whatever you’re looking at. and if it doesn’t make sense at a high level, break it down until you get to you know the most basic or fundamental layer of the data and this will take time but don’t be discouraged, but pay attention to detail, and more importantly, be willing to question your own assumptions and methods.
Be willing to take a critical look at what whatever you’ve been doing, and what assumptions you’ve made, because other people definitely will. One of the hardest things to do is to or rather one of the most common things you’ll have to do is have a to defend your assumptions or defend your conclusions to a bunch of people and so get used to that by questioning yourself. Question your own assumptions, question your own methods. Try and poke holes in your own arguments and then that’ll make it a lot easier when other people inevitably come sort of you know. Come at your conclusions.
Katie: Well, that’s great advice. Thank you so much, Theesh, for coming on my stem career today! Really enjoyed the conversation.
Theesh: Likewise! Thank you so much for having me.
Katie: Thank you Theesh Mohan, Product Reliability Specialist at Meta, for coming on the show today. Listen to every episode of My STEM Career at ellipsiseducation.com or wherever you get your podcasts. See you soon!
Teachers and students: explore STEM careers and discover the ways computer science knowledge can help regardless of your path. In this show, we speak with industry experts that share information about their careers, describe their professional experiences, and offer advice to students. This show is hosted by Codelicious Computer Science Curriculum.