Linguistics at Synchrony (feat. Kyle O’Donnell)
Alumni Aloud Episode 58
Kyle O’Donnell received her Master’s in Linguistics from the Graduate Center. She now works as an Intelligent Virtual Assistant (IVA) Analyst at Synchrony, a consumer financial services company. Kyle also previously worked on Amazon Alexa as a data process expert.
In this episode of Alumni Aloud, Kyle tells us about the power of articulating your transferable skills, the intersection of linguistics and technology, and the advantages of being open to different opportunities.
Listen to the episode below, download it, or stream it in Apple Podcasts (or your preferred podcast player).
VOICE OVER: This is Alumni Aloud, a podcast by Graduate Center students for Graduate Center students. In each episode we talk with a GC graduate about their career path, the ins and outs of their current position, and the career advice they have for students. This series is sponsored by the Graduate Center’s Office of Career Planning & Professional Development.
CARLY BATIST, HOST: Kyle O’Donnell received her Master’s degree from the Linguistics program at the Graduate Center. She now works as an Intelligent Virtual Assistant Analyst at Synchrony, a consumer financial services company. Kyle also previously worked on the Amazon Alexa as a Data Process Expert. In this episode, Kyle tells us about the power of articulating your transferrable skills, the intersection of linguistics and technology, and being open to different opportunities. Could you introduce yourself, what GC program you were in, and what your current position is?
KYLE O’DONNELL, GUEST: My name is Kyle O’Donnell. I was a Master’s student of Theoretical Linguistics at the Graduate Center (GC) with a focus on phonology. And my current title is IVA Analyst, which stands for Intelligent Virtual Assistant, at Synchrony Financial.
BATIST: And how did you get from the GC to there? Can you trace your path from graduation to now?
O’DONNELL: Well it’s been a journey. *laughs* So basically after I graduated, I graduated quite a while ago in 2012 and I really wasn’t sure how I was going to be able to use my Master’s degree or my educational background in a role, something I would be able to get paid for. I kind of dabbled with the idea of going into the PhD program, but thought that track wasn’t really for me. And I will say that while I was attaining my Master’s I struggled to figure out what the practical application would be of something I was spending so much time studying. It created a little bit of tension, well, it created a lot of tension quite frankly. *laughs* Trying to figure out how I was justifying learning about all of this stuff that I loved learning about. So I graduated with not much direction without really knowing what was out there that I could do. I was working retail.
And I was applying to a lot of jobs in various things, trying to get my foot in the door anywhere. And I was in that stage where I kept getting maybe to the 2nd or 3rd interview and then it wouldn’t work out. And so it was kind of a challenging time. And I tend to be a fairly private person and I was trying to come of age with all this technology that lets you put yourself out there quite passively. So at the time I had finally posted my resume somewhere and within days I was contacted by a recruiter and they were hiring quite a few people with a background in linguistics to work on this new project that at the time was at the time quite secretive and mysterious-sounding. And it turned out to be the Amazon Alexa.
BATIST: Wow! That’s pretty cool.
O’DONNELL: Yeah, so I started actually just as a contractor doing some of the less complex work as part of that pipeline and again, I wasn’t fully aware of what was going on behind the scenes; I was working from home. So it was a lot of annotation and transcription and listening to people’s interactions with a robot and a few months later, I was offered to interview for a role, permanently, in-office. And then I became a data associate there, and I was there for about three years working on language analysis. Getting to use my background in semantics and speech processing in ways I really never thought would be possible. I don’t have a background in programming. And I didn’t know that I would be able to find work like that so it was quite a revelation-a very happy one! As I was there I began to progress into more and more of a subject matter expert. I was on a team that was fairly large, full of analysts, and we were all kind of doing similar work on separate areas.
And I think after a few years, though, even though I really loved the work, I was kind of not getting to develop some of the other more general skills that I really wanted to have. And I also just didn’t really know what else was out there. This was really my first full-time role in the private sector and I wasn’t sure I wanted to spend all of it looking at specific types of natural language data. It seemed like a narrow focus. It seemed like, once again, maybe I’m not planning for the long-term. So I left that job and took a job in more general digital marketing. And that was kind of like a very quick activation of a number of things like business skills that I hadn’t really exercised before. After doing that for a little over a year, I was kind of interested to see if there was a way to combine what I had learned through both of those previous experiences into something that incorporates a little more of the broader skills like delivering presentations that are focused on KPIs, not necessarily research and development but delivering some type of return on investment in a very clear way, but also working lots of natural language data which I think is super fun to dive into.
Again, I took a risk in that I hadn’t been on LinkedIn, I had put it off for a long time. Like an impressively long amount of time. And so I created a profile and was like, “Oh wow, I really like this.” It’s kind of nice to see everything I’ve done in one place. And then the recession kind of hit and COVID-19 escalated and things slowed down and I was kind of like re-thinking things again. I just assumed, you know, it would be a while before I would be able to change careers. To be honest I was feeling a little insecure about my current role just because it was dependent on money coming in. And I kind of got lucky again. A recruiter reached out to me and they had this position that sounded like a really good match and I think I started around May.
BATIST: And that’s the position you have now?
O’DONNELL: That is the current position I have right now.
BATIST: Ok. You mentioned you’re an IVA analyst, so you’re working on a chat-bot type thing?
O’DONNELL: Yeah, so on all of their sites they have a chat bot that’s available to help people manage their account, answer frequently asked question or help escalate to a live agent for more complex situations. It’s really cool because they one on every single business page for different partners. So it’s just a ton of natural language data coming in with people typing various things, and I go through it and it’s a lot of chat-bot tuning, looking for language improvements. The team I’m on is actually mostly developers. We’re still in the early phases but eventually the idea is that the analysis kind of leads the development of new features that the chat-bot can do. So it’s kind of heavily tied into the software piece.
BATIST: Gotcha. In a typical day, are you working within a team or do you work fairly independently? What is the kind of dynamic you’re working within in terms of your interactions with other people [at the company]?
O’DONNELL: One of the things I like about it so much is that I kind of have the perspective of the company. I’m on a team of developers who are dispersed around the world so it would be remote regardless of whether or not there was a pandemic. So I get exposure to a lot of different people with different backgrounds. And I’m on stand-up calls at least twice a day with people from my own company but we also partner closely with our AI vendor who provides that service. So I’ll also meet with them on an on-going basis because currently they’re the real subject matter experts on this feature that is so important to the company. So it’s kind of getting two perspectives, one internally and one externally, and working with people from all over.
BATIST: So you had mentioned that you weren’t really a programmer. Did you have to learn those skills through the work? How did you bridge from the linguistics stuff to these more applied coding situations?
O’DONNELL: Yeah. This is definitely one of the more interesting facets of my journey. Because in my mid-twenties, the idea of learning programming was like, “No way, it’s way too late, I could never do this.” Kind of that mindset of, “Oh you know, that window for learning a lot of new things is really passed” was really naïve. When I was working at Amazon, I actually really enjoyed learning to build these little apps that were voice-controlled and that actually motivated me to learn a little bit more about programming, kind of slowly, bit by bit. And then I was also working with people who were good at programming, maybe not full-on developers but engineers who were really familiar with things like Python. They would bring me in very frequently to walk me through their scripts that they were building basically for me and the team’s workflows to use. So that interest just started blossoming over time.
Then, I reached a point where I was kind of looking up things as I needed to learn them when I realized, “It would be great to have foundational knowledge of some of this stuff.” So I took a few online courses that offered through things like Coursera. More recently, since I started this job everything has been remote, and I have a lot more time. There’s no commute or anything like that. I took a class at University of Washington that was really great and to me now this is just something I learn on an on-going basis. One thing I love, which I wish I had realized a lot earlier, is that you don’t have to be a programmer or an expert in any particular language to have it have a huge impact on your skills and what you can apply day to day at work.
BATIST: Did the Linguistics program require any sort of coursework in that kind of thing?
O’DONNELL: So I had the opportunity to take classes like that and I did not take them. It wasn’t built in as requirements and I was just like, “I’ll never be able to do programming. I’ll never do this, so what’s the point in taking this class.” I didn’t see how the puzzle pieces would fit together at that point. And also, I loved the theory. I remember just reading, you know, a few coursebooks beginning to end and being like, “Oh that was great.”
BATIST: I mean I think that’s at least when you know grad school was the right decision, when you reach that level.
O’DONNELL: At the same time, learning the theoretical stuff, learning logic for things like semantics, learning how to approach a set of natural language data from a phonology perspective, have given me so many of the skills I need now for doing the more applied analysis. I think the semantics piece, with the logical formulas, that’s just the one I’ve been constantly so impressed with how useful it’s been and how much it helps me understand programming whenever I return to that.
BATIST: Yeah, for sure. And I think it’s hard to imagine, like oh natural language processing, that’s so big now, that it’s still such a young field, relatively.
O’DONNELL: I remember so few people knowing what linguistics was that I just kind of had to come up with these short lines every time I introduced myself to anyone. Like how to convey what it does and do it justice without suddenly talking their ear off.
BATIST: You mentioned you get to work with people from a lot of different backgrounds. Do they also have graduate school experience? Do you work with other Master’s or PhD grads?
O’DONNELL: The make-up in the role that I’m in right now is pretty diverse and I’m still pretty new there and working remotely so I don’t know if I’m an expert on what everyone’s background is. But it does seem like there’s a pretty good mix of people who pursue more of the business side and program development, business reporting. People with a background in software engineering. And then this pocket of linguists.
BATIST: What about when you were at Amazon? Were you like an outlier in having a graduate degree?
O’DONNELL: Amazon was almost like pulling in all the people with similar backgrounds and putting them in one place to the degree that I was like, “I didn’t know this many linguists existed.”
BATIST: Do you do anything with languages other than English or like translation kind of stuff? Have you gone into that at all or is that something you work with?
O’DONNELL: Yeah so, in my current role now I’m focused on English. Still ramping up on my role here. Previously, at Amazon, we did a lot of UAT testing, so making sure that new areas were working before they were widely released. So basically, before things were released in new languages we would have to do a lot of testing with them. So I was working with a lot of foreign language data.
BATIST: Is it kind of the same principle in terms of analysis or do you have to be fluent in these languages to work with them.
O’DONNELL: Yeah, that’s a good question. Ideally I would say you always have a native speaker working on the language so they can confirm, yes, this sounds right or these are things people would say or they can pick up on those nuances. And so what I was doing was facilitating testing while finding people that spoke different languages to provide that type of validation. So on the one hand, yes, it’s super important to have someone who speaks the language. Within that though, there’s certain things you can do within your role that’ll involve foreign languages that you don’t necessarily have to be an expert on for your relationship to the data, if that makes sense.
BATIST: Yes, yes. You mentioned that you at one point thought about going into the PhD and it seems like a lot of people in the Master’s program do go on to the PhD. Is that something that was kind of like the norm?
O’DONNELL: I think the path for the PhD was a little more well-defined because once you have that PhD, you have the option of staying in academia. Whereas with the Master’s program, it’s less clear what you’re going to go on to do unless you’re pursuing another academic program. So I would say my impression at the time was like “I really love this and I want to do it so I better join the PhD program so I don’t have to stop doing it.” I imagine that’s changed enormously for graduate students and especially ones in linguistics.
BATIST: So what would you say then are the skills that you learned or honed during grad school that helped to get you where you are, that you’re still using today. You kind of touched on a couple, but any other transferrable skills that you found really helpful that your grad career honed?
O’DONNELL: Being able to write about language datasets. I think that’s a really hard thing to do, at least for people who don’t have the background and don’t know how you can kind extract concepts when you’re talking about language and looking for those patterns. There’s a lot of attention paid to write things up [in the Master’s]. I remember some analysis for some assignments and my professor just crossing off a lot and being like, “This is how you do it.” And it was a really great way to learn an effective way to present your findings.
BATIST: Definitely. Do you do a lot of that kind of presentation or writing within your current position? Or is it mainly coding or a mix of both?
O’DONNELL: It is mostly kind of going through giant datasets, finding patterns, trying to combine that with practical solutions that are digestible for a variety of people. So it’s really combining the two. It’s like being able to get deep inside the data and look for these patterns using all these fun new data science tools that are pretty accessible as well as some really interesting ones that our partner has developed because they have a focus on language data. But then turning that into a high-level presentation that you can really convey very quickly and is really easy to understand and shows the value of the work. It’s really important. Maybe I don’t spend as much time on that portion, but it’s probably the most important one.
BATIST: Yes, definitely. Communicating a very complex topic. That keeps coming up in basically every alumnus interview we do! How was the transition when you first graduated from grad school and got onto this Amazon project? How was it transitioning to a more structured company environment?
O’DONNELL: I thought that was actually like an easier transition. For me, being in academia where you have these deadlines that are spaced very far apart and a lot is expected from you and you have a lot of unstructured time. That to me was very stressful. So transitioning to having more of a set schedule with more interaction and more dependencies on other people, I think that went fairly smoothly because I think honestly if you can handle the academia one you can probably do the other.
BATIST: Did you have a specific advisor for your Master’s or was it more dispersed?
O’DONNELL: I was working closely with Julie Blevins. She was my advisor for my Master’s topic and she taught quite a few of the courses I took on phonology.
BATIST: Was she and other professors you had supportive of you going into industry?
O’DONNELL: I think there wasn’t a whole lot of pressure to pursue one or the other for me. It was more just deciding what you want and pursuing it. I think they were pretty supportive. I mean I applied to the PhD track and people were very supportive and then when I kind of broke the news that I actually wanted to take some time off after attaining my Master’s, they were accepting of it and said, “Yeah, that’s reasonable. Lots of people spend some time figuring out what they want to do.
BATIST: Yeah, for sure. You said you were approached at first by someone else at Amazon to first join as a contractor. How did that come about? Was this right when you had joined LinkedIn?
O’DONNELL: So I’ve been approached twice. The first time was for my job at Amazon. I just posted a resume on like Monster or something and it was picked up by like keywords, algorithms. And I was immediately contacted after what felt to me like an endless amount of time just applying to like one-off jobs.
BATIST: When you eventually got the permanent position there, and kind of the other application processes that you had, how was the application process, interviewing like? What was that kind of process like?
O’DONNELL: I think they were pretty standard interview questions so I don’t think there was anything that really surprised me in those questions. And by the way, my experience just in general over the past few years is that a lot of people are really digging into maybe the skills you have not necessarily like, “Oh I had this role where I was previously doing this.” They want to know about different situations you’ve been in and your ability to use different tools, use different skills and also your ability to articulate your own skills. That is really important. But, I mean along with the piece about being able to articulate your skills, sometimes it’s good to just be transparent and say, “This is what I do, this is not what I do.” And I have actually found that recruiters and people doing interviews, they appreciate that a lot.
BATIST: Are there any tips or [pieces of] advice, stuff you wish someone told you when you had just graduated for linguistics students or Master’s students more broadly at the GC?
O’DONNELL: Well, first on that kind of like psychological level, maybe not worrying so much because you’ll find something that’s a good fit for you. And knowing that you have transferrable skills even if you don’t necessarily know them yet. I think it’s really important to not just keep learning things but keep researching how to articulate what you can already do or the little bit more you need to learn to do to make that something transferrable. So that you can talk a common language with the people that do hiring. You get stuck in this feeling of, “Ok, I can talk a really long time about what I do in a lot of detail.” But I just don’t know how to use these buzz words that sound so foreign all of a sudden and that might be obsolete again in a few years. But now I need to say them like three times a day.
BATIST: What are some of those buzz words for you?
O’DONNELL: Maybe not buzz words but working with an agile framework. Knowing what that means, knowing the tools that are associated with it, knowing what the popular tools for doing data analysis are in your field or any field. A lot of times, a little bit of knowledge about how it applies to you can go a long way. You’ve put years into these skills, spend some time figuring out how to talk about them in a way other people can understand.
BATIST, VOICE-OVER: That’s a wrap for this episode of Alumni Aloud. I want to thank O’DONNELL for sharing her experiences in natural language processing. Remember to stay tuned for more episodes of Alumni Aloud, published every 2 weeks during the fall and spring semesters. Subscribe on iTunes and you’ll automatically be notified of new episodes. Also check out our Facebook, Twitter and career planning website at cuny.is/careerplan for more updates from our office or to make appointments with our career counselors. Thanks for listening and see you next time!