Alumni Interviews

Interview with Masa Gumiro, M.S. Data Science class of '18, conducted by Ellie Medina, M.S. Data Science class of '23.

 

Ellie: How did you find out about data science in general and what encouraged you to study data science at Saint Mary’s?

Masa: Well now I’ve been in IT for 21 years. What I've always done is try to learn new skills. Instead of going for full time employment, I opted to do work like contracting. I looked for business opportunities and companies that needed help with technical services. And so I've been in the data space since maybe 2007, 2008. So a lot of time, the easiest way to get more opportunities if you're self employed, is to make sure you're always on the cutting edge. In 2011, I started working for the US Army doing small contracts and supporting them on analytics. And around 2015, I felt like I kind of maxed out my knowledge of that technology. I was looking to see the next step. I did all my research and I kept coming back to data science. It looked like data science was the big brother of Business Intelligence; I kept coming back to data science. So I reached out to my mentor, my best friend, and I said I've got a couple of questions. I'm considering going back to school to do this thing called Data Science, do you know anything about it? At that time, he was in Georgetown, finishing up his Master's degree. So we met for coffee at Starbucks, and a ten-minute coffee plan turned into four hours. And he said, I've been doing data science for the last year. He comes from a Business Intelligence background as well. He said, you need to learn data science. I asked, how did you get in? Well, he was hired by an organization as a Business Intelligence architect. They told him, if you're interested in moving to the data science team, we have some opportunities. 

So around that time, I started researching schools. I applied to about three schools, and I was accepted by all three schools, University of California in San Diego, Southern Methodist University which is in Dallas, Texas, and Saint Mary's. Saint. Mary's was highly recommended to me because the program has a very strong mathematical background. My mentor and I felt like the other programs at that time had more of an emphasis on tools: this data science tool, that visualization tool, that data engineering tool. It seemed like at Saint Mary's their foundation was mathematics. And once you have the core mathematical foundation, then you can learn any tool as long as you know what the engine is actually doing. You can paint the car any color you want. And so that's why I ended up choosing Mary's, because of the mathematical foundation. Also, because of my job, I traveled a lot for the army. So I needed a program where I could take classes online, on demand. I don't have to attend classes at specific times. That's why I chose Saint Mary’s.

 

E:What was your undergraduate major?

M: I went to school at Jacksonville State University of Alabama. I was a computer science major with a minor in mathematics. I started in 1996 and graduated in 2000.

I was offered a job before I graduated. They had a job fair on campus. I used to work in the computer lab, so I didn't go to the job fair. I didn't know anything about it. And they did the whole job fair, they were wrapping up, about to leave. Then the head of department said, there's one other student you need to talk to, he's probably at work. Let me go find him for you, you have to talk to him before you leave. So they called me in and said we need you to talk to these people. I was wearing shorts, I didn't have a suit and tie, I was just coming from work. So I did a quick interview with them and they said, we need you to come and do a technical interview at our location. I borrowed somebody’s car, went over there and did the interview. They offered me a job. I went back to school, finished my last three weeks, graduated, and then started work as a software engineer for a company in Huntsville, Alabama. So I thank my head of department for the connection. Otherwise, I would have never gotten that opportunity.

 

E: What did you do after graduating from the Saint Mary’s Data Science masters program?

M: So once I finished, I’m still in the same job with the US Army that I had when I started the program. Initially, once I finished the program, I went on the job market to see what was out there. One of the challenges I had was that with some of the companies I spoke to, the compensation they were offering was almost half of what I was making at that time. Being the primary breadwinner for my family, I could manage to take a 10%, 20%, even 25% pay cut to get some experience, because once you have experience, you can move up quickly. But if I had to take a 50% pay cut, that's when I would have to sell one of my kids or something to make ends meet. So that was my initial challenge, because a few recruiters I spoke to, even though at that time I had 18 years of experience in IT and at least 11 years experience in data, they saw me as a freshman, a recent college grad. And so that discouraged me a little bit initially. But then my mentor and I also decided to take the business route, the entrepreneurship route. We were like, okay, let's see if we can explore data science projects in Africa. I partnered with my mentor and we set up a company, and we started looking for projects in the US and in Africa. We spoke to one organization in Missouri, an educational organization that was interested in some data science work. We did a proposal for them, submitted it, did the interview with them, but we didn't win that contract. We spoke with an agriculture organization in Africa, we did the proposal, we didn't win that one, either. And then we spoke with another organization in Africa, like a stock exchange, and we actually did win that contract. So we ended up having to do our first contract remotely. We were working at night trying to support them. So that's the route I've taken since then. I'm still working for the US Army, too. The team I work on does not do data science directly. At the moment, they're still trying to do proof of concepts. But you know, we continue to look for other opportunities and projects because I really do want to get more projects, do projects, bring our own people, our own team and run some projects.

 

E: What did you find to be the hardest part about the program?

M: The online environment and not really getting to meet your classmates on a daily basis. But we ended up forming our own little group on WhatsApp or whatever, social media tools. And we're able to stay in touch that way, throughout the course, we helped each other. There were some that came from a math background, but not a programming background. So we were able to see what we could do to help them get up to speed. There are some who came from a programming background, but not as much math, and there were some like me, who have been out of school for 16 years and forgotten all the calculus, but program everyday. So we were just there to help each other and push each other, support each other. Several of us in my cohort had families. So we all understood the struggle. And we were meant to encourage each other and push each other. So the online part was tough. But once I got to know my classmates, you know, it really helped.

E: In what ways has this master's program affected your work in your career today?

M: I'm a senior analytics Business Intelligence analytics consultant. Basically, we do the entire end-to-end data modeling, data extraction, report development, and system maintenance. And then I also do end-user training. I've developed courses for the US Army, I’ll go train them on how to build Business Intelligence reports, maybe two or three courses. And I'm going to teach these courses to army resources who want to build their own reports that's going to teach the actual trainers. But on a daily basis, I am responsible for building data models, within Business Intelligence, data models to support logistics, and supply chain type reports. So I build data models, I do the extraction, I do the performance tuning, I work with the end-user to collect the requirements and the documentation, I do the testing, and then I work with the training team to develop training materials. I do ongoing support for Business Intelligence reports in the logistics area; that's inventory management, material master capacity, planning, acquisition, sales and distribution. I mean, they tie the whole spectrum. So we touch a little bit of everything.

This job isn't really on the data science side. It is more like statistical analytics, which I already had before. I want the next level, which is advanced analytics, AI type models, and so forth. But I did get a lot of good things out of the program that I apply to my job. We take a step back, think of the big picture. We are able to build models that are easily maintainable, can be extended, enhanced, and will be supported. We are not building models that are so specific, like a one-for-one solution, so that if you try to tweak it breaks everything else. I took from the program the whole idea of good design. 

We took a class on data structures and algorithms. Oh my goodness, that was a great class. After I took that class, I had to think about the way I write my code, how to write code that's optimized, and code that operates smarter and more efficiently. As we collect more and more data, we need code that's very efficient, that can process the data quickly, especially as the size of the data increases. So yeah, I take some of those aspects from the Saint Mary's program and apply them to my current job.

E: What has been your favorite part about the program?

M: So coming from Africa, I'd never used a computer. I had no intention of doing computer science. I love math and I’m good at math, really good at mathematics. So when I got to Jacksonville State University, I kind of wanted to go into engineering, but they did not have an engineering program. I thought well, let me do something like physics because I studied physics in high school. So I was wanting to do something quantitative. My academic advisor, after my first semester, said young man, you’re really gifted at math. I think computer science would probably be very good for you, since you love math.  I thought about it, thought let me give it a try. I took a problem class in a language called ADA 95. I was like, this is the best thing in the world. I took another class. I took C programming and I said this is awesome, and then I took C++, and said this is amazing. I just enjoyed the possibility that you can actually tell a computer what to do, that you can write programs and have a computer do something. Coming from where I came from, that was an amazing thing that you could actually write programs for computers, for e-commerce, database websites, all that stuff. So at that point, I changed my major to computer science.

Oh, where do I begin? First of all, I love St. Mary's. I've told all my professors I went to classes with, I love all my professors. They were absolutely amazing. I took a communications class, a project management class, and I took linear algebra classes. I mean, these professors were top notch, they definitely know this stuff. I've been in the industry for a long time, so it takes a lot to impress me, it didn't take much for these guys. It didn't take long for them to impress me. They definitely knew this stuff. The content was rigorous, it wasn't an easy program. I'll be the first one to say that, you definitely have to work hard. For me with a job, family, traveling all the time with kids. I had to juggle everything. I felt like the program was made for people like myself, people working but who want to go back to school to get another degree, it was tailor-made for people like myself. That's one thing I liked about the program. The small class size was an advantage because you could always reach out to your professors anytime, and they were always available. The online concept initially was difficult, because it was my first time doing an online course. I like brick and mortar type classes, sitting at the desk, seeing the teacher, so initially, it was tough, but then the flexibility of the program was just absolutely amazing. So I love the program. I told my communications professor that I'm a tech guy, so when I saw communications in there I was like, why are we doing communications? I need more permanent classes. But once I took the class I said, I see why data scientists should know how to communicate, because you could add these great algorithms that can do all this stuff, but they're not explainable. If people can’t understand the algorithms and the output, then the algorithm is meaningless. So it was a great, well put together program in my opinion.

 

E: What advice would you give to an incoming student in the data science program?

M: When students applied for this program, if they wanted to talk to a former student, sometimes the graduate school would send them my way. And so I've got a chance to talk to a couple of prospective students, and what I've always tried to tell them is, number one, you have  to have the right foundation to take such a program. So first, you know, mathematics; if you went to Calculus that's a good start;, statistics, probability, that's definitely a great start. Because this program, at least when I took it, was heavy on mathematics. When we talk about linear regression, understanding the mathematics behind linear regression with other logistic regression. Neural networks and understanding the mathematics behind it, that element of the gradient descent, Stochastic gradient descent, Batch gradient descent, understanding the math behind all of that. So, having a mathematical foundation helps, you don’t have to be a math genius, but you have to have at least you know, calculus, linear algebra, statistics, probability, those four things I think are more than enough. The other thing is having some programming background. Because what I've seen in the industry, you have to be able to code. There's no way around the code. There are a lot of tools that can do this, drag drop out of the box type features, but once you get in the industry, a lot of the requirements and business cases may require custom programming. That's not part of the typical machine learning libraries. So you might have to code your own classifier, or your own neural network, and so forth. So what I recommend to people is make sure you have had some exposure to computer programming. I think Python is a language that is relatively easy to learn, compared to the old school languages like C, C++, those were headaches to learn. Without the programming it is going to be a challenge with this program or with data science in general. You don’t have to be the best coder in the world but you have to know how to code, what are variables, what is an if-statement, what are loops, what are arrays. I mean, you have to at least know some of those basics and then the good thing is at Saint Marys they teach you from the ground up. It helps to know these things, though. They teach you from the ground up, but maybe at an accelerated pace.

When I took the program, I used to wake up at 2am, 2 or 3am in the morning, then I'll study from 2 or 3am until 6am, and then go about my day. Because one thing I didn't want to do, I didn't want to take away from family time. So from 8pm until 5pm I would be at work, then I'm home. I have to be home; I’m a Dad, I'm a husband, and I spend time with the family. That's the only way I can make it work. This course will require, you know, a lot of work, commitment, but it's absolutely worth it. When you come out on the other side, it's absolutely worth it. I just put data science in my LinkedIn profile. I get phone calls because the opportunities are there. So for students who are interested in it, I was trying to encourage them to go for it. Make sure you have the right foundation. If you don't have it, there are tools to prepare you and to give you the right foundation, because once you get started with class it moves really fast. That's what I recommend.


 

E: What advice would you give someone about pursuing a career in data science?

M: I think it's a great feat. I'm biased because I work in data. I am an IT guy, I'm a nerd, so I'm very biased, but it's a rewarding career without a doubt. If someone is motivated by money, well the money is there. If you are motivated by being able to make an impact in an organization, data science does that, too. You can build models that can help identify new opportunities for companies, identify new customers. My mentor, he built a model and saved the company 1.3 billion. It was just a model. At this company there are robots that will work in an assembly line. And his model was able to identify where the robots are slightly off. And because those robots are slightly off, if a certain robot was slightly off, or configured slightly incorrectly, then whatever the robot did, they had to be redone all over again. So they're spending money having to reprocess some of the materials. And so he built an algorithm and with this algorithm they were able to identify which robots were incorrectly calibrated, and how to calibrate them correctly. By calibrating correctly they were able to save about 1.3 billion, because now they don't have to reprocess all that stuff. So data science has these opportunities where you can make an impact on an organization just by writing the code in the comfort of your cubicle, or your home, and so forth. So  I'm all about data science, I'm all about data. If anybody is interested I would say go for it. 

When I did my undergrad the hardest thing was knowing the next step, what to do next. That was the hardest thing. When you are in school the most important thing is a recruiter, someone that can help you get a job. To take your skills, match them to a requirement, and get you a job. What's important is a mentor who can see the work you've been doing, your education, your experience, and then someone who can say, based on your experience, I can see you one day being a Chief Technology Officer for a company. For example, someone who says, I can see you being a director of analytics, and here's what you need to do to get there. These are the steps you can follow. Why don't you go do a Master's degree? Why don't you take these graduate courses? Why don't you take these, you know, PMP certifications. So these are the tools you need to qualify to be a director of analytics, and so forth, and you need recruiters to help you get a job. But it'd be nice, even before you get to mid-career, if you had a mentor, someone that can say, here's where the market is going. 

Maybe this data science thing is the right way to go. Access to information used to be so difficult, but now with social media, and all these TED Talks, and YouTube, information is readily available. But you still have to process it to get all that information. For people who are into data science, go out there and do some research on the internet. There's so much information out there, there are forums and great discussions. There are websites with paid services where you talk to a mentor, and they can direct you and so forth. But if you come to Saint Mary's, you get some of the best people who can help direct you with where the future is, the trends out there. I love the professors at Saint Mary's. 

I remember when we did our orientation and one of people on the board was a gentleman who works for Netflix. He was on the advisory board for Saint Mary's Data Science program. After we graduated, we got an email that Netflix was hiring, and if anyone was interested to reach out to him. Through Saint Mary’s own network, we're able to get connected to other people. And even if I'm on LinkedIn, sometimes I'll see a job, take a look at it and see a flag that says, somebody from Saint Mary's works for this company. Data science is a great field. Anybody who is interested, please pursue it. There are more opportunities than people out there. So even if you may not be genius level, there's enough opportunity for everybody out there. And it's a growing field. I know it seems like people say data science has been there for a while, maybe it has, but being accepted universally, that has happened within the last five or 10 years or so. The techniques have been used for a long time but the technology we have now, that allows us to process petabytes of data very easily, the computing power we have with cloud computing, and data size now being more readily accepted all over. So the opportunities are there, I think that field will generally grow. As time passes, I think there will be more and more opportunities, as more companies begin to adopt data science as a key resource as a need to drive the companies forward.

 

E: What are the projects you do outside of work like?

M: The last project we did was a combination of data engineering, data science and Business Intelligence. So it had those three parts. Basically they were implementing a data warehouse for the first time. So they could bring all data from all the business units into one repository, and then build Business Intelligence reports on that. So they had somebody do all the work and then towards the end, there was an issue. So they brought us in to help with the ETL with bringing the data, the Business Intelligence reports. And then the last part was the data science part. Now for this project, because of their funding, they only had money for the first two pieces. So we did work for them, we did analysis of what was developed, and then we started development of the actual Business Intelligence reports. At some point, they probably need some re-engineering of some of the work that has already been done. And so that's the extent of the work we did for them. We never got to the data science part because they didn't have the budget for it at that time. So we ended up just doing the data engineering and the Business Intelligence project.



 

E: What are your hopes for the future?

M: So my friend and I came up with an idea to build our own out-of-the-box product. A machine learning product specifically for the supply chain industry. So we have been working on that, brainstorming about it. Because what we see in the industry now is a lot of people want data science and a lot of people want machine learning. But they don't know where to begin, where to start. If you go into accounting, for example, it is assumed that if you go to insurance companies, there are certain models that are known across all insurance companies. If you go into accounting, there's an income statement report, it’s the same across all companies, there's a balance sheet, etc. So we thought, why can't we build a kind of plug and play solution that has pre built data science models specifically for supply chain metrics. That's what we're trying to see, how we can build data science models. Then if a company wants to use them, have API that will ingest, stage, clean, process, and then feed the data into the model. Then they start to get the benefit of data science. So that's a concept we've been brainstorming on. So that's one thing I'm doing. 

The other thing is, I'm always learning. I started a Google class. So Google has an amazing data science library called TensorFlow. Over the years I've kind of been hearing about it. So I finally said, You know what, I want to go get certified as a Google Developer through Google Cloud’s Data Engineering Certification. So I actually signed up for some courses I'm taking online right now, so I can become a certified Google Data engineer. And now I'm going to be a certified Google machine learning engineer, as well as a certified Amazon machine learning engineer. Those are three courses I'm taking right now just to learn their specific tools and so forth. Because in the cloud world everyone is heavily dependent on cloud solutions, whether it's Google or Microsoft Agile or Amazon, Amazon Web Services (AWS). So I'm always in to learn the latest technologies, latest tools, and adapt. I'm definitely exploring to see what's out there. As far as data science data engineering roles, I'm definitely interested in finding more opportunities in those areas.

I’ve been in IT for a long time and what's allowed me to keep getting opportunities as a self employed person is whenever there are new technologies, I will quickly go learn them, try to get certified, master it. Then I'm going to a client and I can say, if you need help in this area, I'm certified in that area, I'm an expert and can help you. So to stay ahead like that, to stay marketable, you always have to stay on the cutting edge. And now access to information is much easier now. Back then it was hard to find a trainer, but now you go on Coursera, you can go on Udemy, you can go online. There are so many courses on anything, they're affordable, they're right at your fingertips, it’s just a matter of being willing to spend the time and dig out an hour each day just to study, and so forth. Because what I've done is an hour a day, and so right now I'm learning Google. After that, I'm learning Amazon Web Services (AWS), some of their latest tools. I've started looking into autonomous vehicles, self driving vehicles, with computer vision. And I'm interested in exploring that area as well. How do we use self-driving vehicles? I'm very interested in the Internet of Things (IoT). I did my capstone project, I did the IoT projects. I'm very interested in this internet of things where we are collecting data from all over the place and making sense of that data, and using data for decision making. So I am very interested in exploring IoT, particularly for agriculture, and maybe the use of drones, flying around collecting data, and then using that data to advise farmers on the type of crops and the type of seeds that are good for their particular location, their environment and so forth. So I definitely have a lot of things I'm interested in pursuing. You have to keep learning otherwise, with technology, you'll be outdated very quickly.