Resources
Feb 23, 2022 |
1h 2min
|
1.4k views
|

Data Science Hangout | Katie Schafer, Beam Dental | Building a Data Science Portfolio

We were recently joined by Katie Schafer, VP of Advanced Analytics at Beam Dental, to discuss the most important things going on with data science leadership.

Great tips & resources shared for this question: What can I do to be able to present my skills and build a portfolio if I’m not able to share my work?

➕Many people expressed that they are in this position.

Katie: Is it possible to simulate something akin to some of the data you would use? For project portfolios, submitting the actual work is not something that I’ve seen often.

In interviews, you need to be able to discuss the work. Describing the projects is really tough because you want to say something meaningful but you don’t want to be too specific. Having a role play conversation with yourself on questions you’re getting & practicing answers that are more generalized can be helpful. Take one specific problem, translate that to an abstract problem that you tackled with statistics and programming.

Javier: I’ve been in a similar boat. There’s a package called synthpop: https://lnkd.in/g_paFGNB, which helps take actual data points and simulates them to make a completely synthetic data set.

When I was interviewing for my job now I built a shiny app that I deployed to a free shinyapps.io. I styled the whole app to be consistent with their brand. I don’t attribute that alone to getting the job, but the interviewers were super impressed. The ability to change out the theme is so easy, so for anyone who’s trying to get a job this could be an easy way for you to have something that you can change whenever you interview with a new company.

Rick: If you’ve ever worked with a designer - they have a template and they’re not just creating everything from scratch. If you have a nicely organized project as a data scientist, you’re showing your skills as a software engineer.

Zac: I do a lot of personal projects because I don't use R at work. For these projects I focus the topic on what skills I want to show off. I’m looking to start my career, so I write a blog & include reasons why I did something. I think this makes it a lot easier to share what I’ve done https://lnkd.in/gzhwuCqw

Libby: For me with portfolio work, it's been difficult because sometimes my contributions are one algorithm inside a larger shared project, so I stick to descriptions, talking about what I did and why I did it, what the result was.

Larry: I describe the topic and present the deck. Sometimes I just make fake data, but talk about the steps I took.

Katie: Some of my richest connections have also come from speaking. R-Ladies has been a great starting point for that. A lot of meetups have gone virtual so you can now join a chapter even if you don't have one in your city.

Jiwan: Maybe start with a 5 min lightning talk with a topic you're excited about!

Other resources shared during the chat:
Prabha: Technical debt in ML systems - https://proceedings.neurips.cc/paper/2015/file/86df7dcfd896fcaf2674f757a2463eba-Paper.pdf
Rachael: R community explorer app: https://benubah.github.io/r-community-explorer/rugs.html
Javier: GitHub for the {bslib} app, feel free to clone as you see fit: https://github.com/JavOrraca/bslib_demo_shiny

Jobs shared during the chat:
Niklas: Several data science roles at my company (sustainability SaaS). 2 data engineers, 1 data analyst, 1 data scientist - https://higg.com/careers/
Jill: We are hiring for a Data Analyst role here at Viemed. We are a healthcare company and we LOVEEE DATA. https://viemed.apscareerportal.com/jobs/1338665/apps/new?embed=1"

image: thumbnail.jpg

Transcript#

This transcript was generated automatically and may contain errors.

Welcome back to the Data Science Hangout. I hope everyone's having a great week. If you're joining for the first time, welcome. It's great to meet you. I'm Rachel Dempsey and I'm the host of this Hangout. Basically what it is, it's an open space for the whole data science community to connect and chat about data science leadership, questions you're facing and what's going on in the world of data science.

If you ever want to go back and rewatch prior sessions, or share with someone who's missed it, we do have a brand new data science Hangout site. I think I can still call this new for a few more weeks. But Rob can share that in the zoom chat right now as well. We really want this to be a space where everybody can participate and we can hear from everyone. So there's three ways you can ask questions. You can jump in live, raise your hand on zoom might be the best way with a bigger group. You can put questions in the zoom chat. And feel free to put a little star after your question too if you want me to read it. Maybe you're in a coffee shop or something. And then we also have a Slido link where you can ask questions anonymously too. And we'll share that in the chat window as well.

We really love the dialogue that happens live during these Hangouts. So if you do find yourself inspired, and maybe have spoken a few times, maybe consider holding back a bit for others to jump into. So we want this to be as inclusive as possible. And making room for others is also an important part of that. But again, just want to reiterate that we love to hear from everyone, no matter your level of experience or area of work. But with that, I'm so happy to be joined by my co host for today, Katie Schaefer. And Katie, I'd love to turn it over to you to just introduce yourself and maybe share a bit about the work that you do on your team.

Yeah, absolutely. Well, hi, guys. It's great to be here. I don't know if anybody else is in the Midwest region, but heavy storm here. So hopefully you guys are all staying warm if you are. So my name is Katie Schaefer. I work at a company called Beam Dental. So we're an insurance company and we focus on ancillary benefits. Some work that we do on our team. So I oversee the data science and the analytics team. And this was kind of my first big focus on an analytics team. So it's been really fun to kind of see the differences and skill sets and focuses between the two teams.

And we do a huge variety of things across the two teams. So on the data science front, most of what we do is kind of the more traditional predictive modeling problems, getting those into production. And most of those are business facing. So we do pair with our underwriting and actuarial team. But they own kind of those traditional actuarial models. And we handle a lot of the business facing data science problems on the analytics team. We do all sorts of things. So a lot of analyses, but also a lot of what I like to call data consulting. So folks from the business coming to us with key problems that can be answered or eliminated with data and us working to try to provide them some additional information, some opportunities for process improvement.

And from a tool stack perspective, our Python, Python's pretty big on the data science team. And also Looker, we recently moved over to Looker, which is exciting from Tableau. So lots of SQL too.

Nice. So I know there'll be a lot of questions coming in. But to kind of open up the discussion, I know, in the hiring panel that you were part of, hiring great data science teams, we talked a bit about the involvement of in the open source community, and how that impacts hireability. And I think it'd be interesting to just kind of dive deeper into your perspective on on that, too.

Yeah. So I can just give kind of a general overview, and then excited to kind of dialogue from there. So I think it is kind of rarer than I might have expected that I see some heavy involvement in the open source community on applicants. So when I do, I get pretty excited for a few reasons. So I think from my perspective, it shows a true kind of self driven desire for learning for improvement, and for kind of staying fresh on all the tools within the industry, and latest packages, etc. So it speaks to that kind of genuine interest in the job you're applying for versus versus just kind of a nine to five that somebody's focused on.

In my perspective, I think some kind of watchouts there is obviously there's a huge range of involvement in open source communities. So I think really what's super rich is when I can dialogue with somebody and hear how amazing their experience has been in that open source community and that data science community. Because that to me really speaks to kind of their passion for the topic and the job at hand. So that's kind of a high level how I like to think of it. It's always a nice, a nice thing to see.

And I'd love to kind of learn more about it. Awesome. Yeah, I'm sure there'll be some follow up questions on that too. And it'd be great to chat more about it. But also, maybe to help us all think like put ourselves into the mindset of the work that you do. It'd be great to understand like, what are a few like questions your data science team is solving or business problems that you're helping solve as well?

Yeah. So from a data science perspective, at a kind of super high level, one problem that comes across multiple different domains of our business would be kind of at a high level entity matching. So you can think of this as a common problem in any business, right? But when individual entity comes in, you've got kind of name and address, how do you pair it with the names and addresses in your system appropriately and at scale. So that kind of hits a few different places that being whether it be new business and leads or more on the kind of the claims aspect of things. So that's kind of one current project that should give you a bit of a sample.

And then also kind of on the business facing front, always an appetite for knowing information sooner than it's available. So if you think of an opportunity coming in, knowing how much that opportunity is going to end up at from a dollar perspective, knowing which opportunities were more likely to win or which clients, you know, kind of have the optimal potential from a cross seller upsell standpoint. So some kind of really general business problems in the insurance space. So that's kind of a high level overview of from a data science side. I think some of the things that are top of mind now.

Cool. And when you say like when those opportunities, what does that actually mean for Beam Dental? Yeah. So for Beam, we sell insurance. And so we sell as any kind of healthcare insurance typically works. Usually we work with brokers who work at brokerages. So it's kind of a B2B2C space. So it's not like a car insurance company where you sell directly to the consumer. So there is kind of a sense of how can we prioritize the service for, you know, our swath of brokerages we work with. So probably similar in kind of a direct to consumer space. You know, who do I send my really high touch marketing to? Who do I spend a lot of time doing outreach to?

Open source community and Stack Overflow

Yeah. Okay. Thank you. That's super helpful. Yeah. I see there was an anonymous question that came in on Slido about talking about open source. So when you think about that, how does Stack Overflow fit into that?

So I am maybe a bad example of a Stack Overflow contributor. So I am very much a consumer of Stack Overflow. I'm a little intimidated to my own questions, but I think Stack Overflow is hugely kind of, obviously it's impactful. I'm sure every single person on this call has used Stack Overflow quite a bit. I have never come across somebody who was a big Stack Overflow contributor from an interviewing perspective, but that would be amazing if I did. I think actually one of my coworkers' father is one of the great contributors to Stack Overflow and our forums. So I know I was just super excited to talk to him. And to me, that's just the highest honor in terms of contribution to the open source community, in my opinion, is really contributing to Stack Overflow questions, reproducible examples, but then the answers as well.

And to me, that's just the highest honor in terms of contribution to the open source community, in my opinion, is really contributing to Stack Overflow questions, reproducible examples, but then the answers as well.

Analytics vs. data science

Awesome. I see Tatiana in the chat. You had a great question there and I'd love to have you ask that live too. Sure. I was about to type in. Sorry.

Yeah, sure. It's more of a philosophical question to get us started. I imagine you have one team, obviously, but since you do present yourself as a director of analytics and data science, I would be curious to hear about how do you think about the differences between the two? And I see that initially you were a manager of advanced analytics and then data science was kind of like joined afterwards. Does it mean that you started undertaking more forecasting and predictive analysis and how do you think about the two? And if you have the split in responsibilities in your team members who are responsible, one for advanced analytics and one for data science, how you can comment on that?

Yeah, that's a great question. So I think the delineation of data science and analytics looks probably a little different at every business. So I'll just caveat it with my answer is not the right one, in my opinion. It's just kind of how I think about it. So data science and analytics are two different but really related teams. So they work together quite collaboratively.

So we have shared lab meetings, shared team meetings for now. I think as teams grow in size, obviously, that can be harder to keep that cohesion together. We also really work extremely closely with data management now. So we have, you know, same stand-ups. I could see in the future there'd be kind of from a just like functional perspective delineations of that from a meeting perspective. But I think it's really important to keep that bridge together.

In terms of how I think of the differences between analytics and data science. So first thing that comes to mind is from a tool stack perspective. Traditionally, I think the analysts are much heavier on the SQL skills, much heavier on BI skills. So a traditional analyst that I'm typically interviewing or hiring is much more likely to have experience with like Tableau or Looker versus a Shiny. Shiny, for example, typically when I start to see example with something like Shiny, I also see really strong programming in R or Python. And then often it tends to kind of correlate with the deeper statistics background. So that's from a tooling perspective, some delineation. So our data science team definitely more focused on Python and R versus traditional BI platforms. SQL really common amongst both of the teams, but it's pretty strong among our analysts from a skillset perspective.

In terms of functional problems, how I like to describe it is I view the analyst as kind of the business data consultants. So they're often tackling kind of vague questions. So it might be a trend that's being observed in the business and their task is to kind of dig in from a data perspective and find indicators that might be indicative of why that trend is occurring. And also maybe qualitatively, or I should say kind of from a trend perspective, indicative of what might happen in the future. But they might actually not get to the level of building the predictive model to make that happen. It's more kind of extrapolation from visual trends often, and some basic kind of descriptive statistics. And then a heavy amount of like data storytelling, which I know can be a general term, but I do think it's a core aspect of what they do. And then also just bread and butter reporting. So how are we doing as a business on our key metrics, right? So wins, our deal sizes, et cetera, just the staple daily refreshing tracking of that.

Data scientists, often they will be working on a very specific problem that was uncovered from the analyst. So usually a predictive model. There are definitely some automation projects that are using tools like Python, but in my mind, kind of more traditional automation work. So if you could imagine something like extracting information from PDFs, right? Requires kind of a data science skillset, might be a probabilistic model underneath that, but could also just be solved with some kind of classic tools. And then from an end user's perspective, you are automating their workflow, right?

Even though it is in essence, just programming. So data science, more specific focus problems, less stakeholder facing time, more development time, predictive modeling problems and automation analyst really partners to the business and those data consultants. So hopefully that helped. That's just how I have the view I have shaped, but could be much different other places.

That's fantastic. Thank you. Yeah.

Managing a mix of R and Python

Thanks Katie. Mark, I see you had a great question in the chat too, if you want to jump in.

Yeah, sure. Sorry, I was on mute. Katie, you mentioned you have your data scientists on the team use both Python and R. I guess like how do you manage that mix of languages within the team?

That's a good question. So admittedly, I am the person who is the R user and R2 kind of core data scientist or Python users. So I'd say the vast majority of the development code and production code is in Python. In terms of how we manage it though, I think it seems simpler in practice than I pictured it to be. So both of the data scientists are really strong in Python and have dabbled in R at least, if not used it in a master's program, et cetera. So they can read R and I can read Python and vice versa. But if either one of us were going to build something, it would be much quicker to do it in our preferred language. I am kind of a proponent of the right tool for the job. So most of these things we have not tackled quite yet, but on our roadmap, there are certainly problems that I foresee R being the better tool because there's a more robust set of statistical packages out there for kind of niche problems.

So I'm a big proponent of right tool for the job, right tool for you, but I'd say anyone who's fairly strong in one typically doesn't have problems reading the other or interpreting the other. So I will often do code reviews or coaching sessions on Python code, even though I'm not the expert in Python, it's pretty quick for us to get on the same page of what I'm trying to convey or what they're trying to convey. And conceptually, most of our like working sessions are going to be more about the model we're choosing, the performance metrics, et cetera, features that we want to include, and a little bit less about the specifics of code. So if you're hiring kind of really new novice talent, yes, absolutely. I do think it's critical that you're really strong in the language you want to train them on or somebody who's senior

on your team is, that would be a problem. But since we kind of started out with some more senior folks, they are pretty savvy in one or the other, but they do have a preference.

Cool. Thank you. Yeah.

Thanks, Katie. Kind of following up on that, I see Robert had asked the question as well, that was, how do you think through which tool is the right one for the project or requirement?

And I think that might go even into like when you talked about the data analyst side of the house using Looker, for example, too. That's a good question. We actually have kind of an example I'm thinking of right now where I'm trying to guide kind of our senior data analyst and our data scientist on which tool to use. So I usually start with the end user of whatever we're outputting. So, I mean, if it's a model, it's kind of a no-brainer, like Looker, the BI tool, SQL, probably not the right tool set. So for more of those problems where it's kind of automation or the model is being used as an input somewhere that may be more business-facing and less of like a number where things are being calculated off, that's when I feel like that decision comes into play.

So right now we have a bit of an automation project, basically at a high level, just automated reports that need to get churned out every month with a heavy amount of math and calculations under the hood. So right now I'm trying to evaluate for the end users of this, what are their primary needs? Do we need this to be static, auditable, and traceable? That might lend more to Python or R Markdown, right, as the tool set. Or do they need to interact with elements of this and there's less of that need for it to be a static PDF?

That might put me more in the Looker camp, right? So there are aspects of it where they want to kind of see how they're doing internally for our sales team monthly, where there's just the static value could work, but they also kind of want the ability to like input new values and see how those change or how much basically something would be worth to them when they're kind of prospecting it.

So without going into too much detail, that kind of brings me down the path of like, wait a second, what tools are really required to meet that end user need? So I usually start with the end user and then back up from there and that's kind of the same process the team goes through.

I like that a lot, starting with the end user. Yeah. Jill, I see your question you said was answered with that as well, but I think it brings up a point to me to think about, like you have the data analysts and data scientists as well and they're using different tools, but do you sometimes have people moving from one side to the other, like data analysts starting to use R and Python and then- Absolutely, yeah. So we definitely have analysts who are more senior ones who are fairly strong in Python, but they're definitely stronger in the BI tools and SQL.

So there are examples in the past before we had a more built-out data science team where they were doing some of that automation work and they were doing some like API calls for automated sends and marketing material, for example. But definitely I try to think of like, where is their skill set strongest? And it's going to take them longer to figure out how to do something in Python than it will if they just are using the tools that they use every single day. So now that we have that capability, those sorts of problems are probably no longer going to them, but they can interface with that.

I would say from a data munging perspective, it has been super illuminating to work with like a pretty advanced analytics team to see how much they do in SQL in terms of data manipulation. It is pretty impressive.

Being more of an R user, I honestly wouldn't think of that as my first choice for things like pivoting, et cetera, but they are pretty strong in SQL. So they will often use that. That being said, there are some of them who prefer to do that sort of thing in like Python. So they'll just kind of query the database, do their data manipulation in there, and then kind of write back if they need to.

Okay, thank you. I'm just scrolling through to see which other questions we didn't get to yet. And Zach, I see you had a question in the chat too. Would you want to jump in and add some context on that one? Yeah, you mainly asked my question, Rachel, about that. Yeah, it was just a case of, it's more like, it also goes into the internal politics.

Are the data science people seen on the same level as analysts, or are the data science people seen as above? Just like career and training, because we quite often talk about training and how to upskill people. I was wondering, given that you have used Looker and Shiny, try and promote people to use Shiny instead and upskill themselves, and then move to become a data scientist? Or is it just, you specialize in this, this is what you're best at, use this?

But again, you mainly answered the question.

Looker vs Shiny and tool strategy

There's a good question on Looker and Shiny though. So to clarify, nobody today, at least in my current company, is using Shiny. That was more of my experience. So I was trying to say, BI to me was Shiny kind of before I was more exposed to the analytic side. So that would absolutely be kind of an interesting intercompany trade-off if people were using. I feel like I've talked to people at companies who had like Mode, Shiny, Tableau, like multiple different ones, right? So that is kind of an interesting question on like, who wins? Where do the resources go? Do you focus on moving people to Shiny if they are using Looker?

I'd say the way I would personally think about it is, what is the quickest, on the analytic side, what would be the most efficient path for getting the business what they need? And if my analysts aren't super strong in Shiny, that probably is not, to be honest, the tool we're going to use versus my data scientists, right? Like if they, so right now, obviously their main tool is Python, but I would never ask them to learn Looker if they are comfortable building a web app in Python, like that wouldn't make any sense. And the same with an R user. So I think definitely in kind of our growing product analytics side, that seems like it kind of draws more of an R crowd. So I could absolutely see them doing things like churning out a Shiny dashboard. And I think the organization is at a size in which we're kind of serving different pockets of the org. So while you don't want to kind of increase technical debt, I do think it's possible that like a certain group of end users might have more of an inclination to kind of the look and end feel of Shiny versus Looker.

And so I would never have somebody kind of move over to a tool they're uncomfortable with, unless there was a good reason for them to do that. Okay, yeah. So, but you would try and help promote, like, so you would just say this, do what's most comfortable, rather than saying, we'd like you to eventually move to Shiny, for example. Unless there's a clear business need to then yeah, absolutely. I would not just stick with what's comfortable, right? So I guess, different than R, but Tableau and Looker are good examples of that, right? Like, most people we'd hire, most people on the team are more comfortable in Tableau. But from a use case perspective, the customizability of Looker just seemed like the way we needed to go from a business for a lot of the use cases we have. So while Tableau is the comfortable choice, we're still going to transition over to Looker. And I would hope the talent that we have is always going to be on board for learning something new, so. Okay, yeah, thank you. Sorry, I don't know the difference between Tableau and Looker much, so I was trying to use examples. Oh, no worries.

Hiring for data science roles

Yeah. Thanks, Katie. I saw, I think, maybe on LinkedIn at some point, that you were hiring for data science roles a few months back, and was just curious to dive deeper into, like, that topic of hiring too, and, like, what are particular things that you look for in hiring on your own team? Yeah, so from a data science perspective, I mean, the bread and butter staples are that you can program. And I do kind of primarily only focus on R or Python. So while I would absolutely consider a candidate who is really strong in SAS and maybe had never used R or Python, I just feel like the talent pool, there's just, it's just already so rich in one or the other that I would not really, unless there was something else really shining about that application, I would definitely favor somebody who has those skills in R or Python.

In terms of which one or the other, I'm kind of agnostic, and that might be because I come from being more of an R user, but I think it's just hard enough to find good people. I think trying to put that added layer of, like, you have to be strong in R or Python is way too restrictive.

So I think the other thing, just based on where we are as a business, is the data scientist candidates I was looking for definitely needed some kind of element of not just developing the model, but also scoping it, delivering it back to the business. So we certainly lean heavily on myself and the analytics team for that portion of things, but it was also important to me that the candidate was kind of senior enough to have that exposure.

And yeah, that's helpful, too. I'm just thinking more about, like, how you found candidates as well, because I know, like, a lot of people have been sharing job postings and saying it's difficult hiring right now, and where, yeah, where do you generally find candidates?

Yeah, I mean, to be honest, it took a while, right? I mean, maybe it's the planning fallacy, but I thought it would take a while, and it took longer. In terms of where we are, in terms of where we found them, so one ended up being a referral of the other, which is kind of the best situation, but I believe kind of it was just LinkedIn seeing the job posting, but I absolutely got some great candidates who just ended up not being a fit for a variety of reasons through open source community forums. So I posted on Our Ladies. I posted on kind of our local Columbus tech org channel, and I definitely talked to amazing people from those sources, so that's kind of always a first go-to for me is, yes, our recruiter puts it on LinkedIn, but then I take that link, and I push that out to all the open source communities that I'm involved in, so. Awesome. Thank you. Edgar, I see you had your hand raised. You want to jump in?

Transitioning between industries as a candidate

Yeah, so I have a question. So when you're looking at candidates, and someone's looking to transition from one industry into another, how do you take that into account when they may not have the most relevant experience in the industry that they're in? Yeah, there are, I'm trying to think of, like, of our whole team, there's probably only one or two people who had insurance experience.

Obviously, I favor that if you had insurance experience, but again, I just find that it's pretty hard to find really good people who fit your specific needs and the time you're at an organization. So there are a ton of amazing data scientist candidates out there, but starting out fresh, it wasn't really the best play for me to think of somebody super junior, because there were just, there wasn't enough, like, coaching time to go around, right? So I know that can be a mixed message, because you might be thinking, like, there's tons of great candidates out there. Totally are, but I think when you're in kind of an early to mid-stage growth company, your role requirements get a little more narrow, and it just can be hard to find really good people. So I, hopefully, that kind of answered your question. Is there any kind of portions I didn't address from that?

I think it's, yeah, I think also kind of making yourself competitive, I think, in that terms, and you did answer that to a degree, but I think, like, I'm transitioning from education into, like, a different industry, and so, you know, how do you get past that initial, where a recruiter might say, how can I make these connections from one industry to another? And I think...

That's a great question. So, like I said, we definitely have hired people without insurance backgrounds, but I think what has made them really attractive is they have some sort of demonstrated ability to learn, right? And showing a passion for that, right? So I can teach you the industry. It's much harder for me to teach you, like, foundational statistics principles, to teach you coding and programming. So that's always my approach, but I will say what can be a little challenging is if you ask somebody for examples of when they taught themselves a new concept, or learned a new process in the past, and they come up blank, that can be kind of a red flag for, like, hmm, like, will they pick up on the business aspect of this as quickly?

I know coming out of academia myself, that can be a tough transition, because you, in reality, if you've got a PhD, right, like, you have four to five years of job experience, but people don't often look at you that way in the market, unfortunately, and I think that's a big miss. Unfortunately there, I think a lot of that is you just getting the opportunity to sell yourself, right? So, like, if you can just get that first interview and demonstrate how the skills you learn in your academic position will translate, and, I mean, what a better example of, like, grit than completing a PhD, right? Like, if you can sell yourself on the transition of your skills, I think you can succeed in that interview process, but I think where people probably, unfortunately, get overlooked is just getting that first interview.

So I can teach you the industry. It's much harder for me to teach you, like, foundational statistics principles, to teach you coding and programming.

So I think a lot of that, if you're coming straight from academia, is, like, tailoring your interest to the right companies. So this is just a generalization. I wouldn't say this is true of every major company. Absolutely not. I've definitely seen large companies that have a appreciation for academic talent, but in general, I think startups or mid, like, mid-stage growth companies tend to be a little bit more friendly to the academic crowd, because there are folks there. I think if you can find a manager that you might be working under who has their PhD, they are going to be more alert to you as a high talent potential.

So our recruiters, for example, they are the first stop, but I am looking in there all the time at people coming in, and I will tell them, hey, just because this is a senior role, don't roll somebody out just because they have a PhD and no work experience. I want to see them, but if you don't have the hiring manager doing that, then you might get overlooked, and you might not even get that first conversation. Thank you. Yeah, Katie, and actually, like, elaborating on that a little bit, I can see that Tatiana asked a great question. If you want to jump in, Tatiana,

about how Katie went from one area to another, moving into analytics. Yeah, that was very interesting, and I recently interviewed for, or actually, I was interviewed for how I shifted my careers, and one of the aspects that I highlighted was self-promotion, especially we as, I don't want to, like, generalize or miss everybody out, but as women, we tend to undersell ourselves, I feel like, and it's very important not to, like, get all that mental self-talk out of the picture and just go ahead and sell yourself, because if you don't, if you don't speak up, and if you don't know how to promote your persona, how to think through the eyes of the person who is hiring you, meaning so you can showcase how your analytical thinking that you developed during your PhD or your master's, or whatever, how it translates into your business case is very important, and then you can do projects on the side that actually show that here is an example of how I did it, because nobody prevents you from doing courses on the side and do projects on the side, but this is my penchant, and the question was if you could speak more about your personal experience, how you transitioned from psychology to analytics and data science.

Yeah, so I was in a PhD program, and they can vary, but mine started in clinical psych, so there's a heavy focus on research, so Ohio State's a pretty big research institution, so what was clinical psych, and so technically a part of it is getting your license as a clinician and seeing clients, that was not the intent of my program, so like, my professor hired me because he, he chose me because he wanted me to work on academic research with him, so that was a heavy amount in training of statistics, statistics courses, experimental methods, and then programming as well, so that alone though, especially in today's market, would have been a bit hard to sell as a transition, and I feel like what you need to like, focus on is that one, that first opportunity that's transitional, that's the bridge, I think you can get it from user groups, and that's another huge value of the open source community. For me, it was a kind of graduate assistant social associateship with Nationwide Insurance, so they basically bought, you know, kind of fledgling PhDs out of their teaching contract, which they didn't want to do anyways, and they, you know, would pay them to work in their analytics department, so it was a great gig for me, and it was that like first exposure, and then once I had that, like my network opened up, I had quote-unquote industry experience, even though it was white, and the path was just open from there for me to fly on.

I think with masters of analytics and data science programs today, it's like a little, I think some of these core PhD and master's programs can get overlooked, and I think the way to sell yourself is to, one, just learn to sell yourself, which is a very foreign concept in academia, but two, also to get that first experience, and I think if it's hard to find, user groups are a great way to start. Open source communities are an exceptional way to start, because you will, at the very least, dialogue and meet people who can expose you to that first industry opportunity that's more formal than the, like just user group itself. I will also say there's a few programs, so I'll throw one in the chat, that I am like volunteer at, so one is called the Erdos Institute, and it's founded by math PhDs, and their whole purpose is to train PhDs who have decided academia is not for them to get that industry experience, so it's kind of like a boot camp, but the talent is just incredible there, and also the training and prep for industry, so that's kind of another like segue example, but like you kind of just need to find that first connection, I think, and that's the hardest, hardest part.

Skills and hiring for entry-level data scientists

That would be great if you could share those links after, Katie. Yeah, I can share them on LinkedIn too, but kind of following up on this topic too, there was an anonymous question sent to me that was, what are the skills that you look for in an entry level data scientist from someone who has a degree in operations research and financial engineering, but is having a difficult time landing a job, and would love to just hear more about like what tips you have for selling like yourself and getting that role too. Yeah, so entry level, I'd say the things that I would look for is, I guess it's all basic, and I'm probably talking, or all relative, and I'm talking to some probably fairly advanced R users, but at least like that basic foundational skill set, so we have a take-home project as part of our interview process. Also, I think in just asking questions as an interview, you can tell how deep their skill set goes, but from a basic level, I would hope they could do the bulk of the analyses in either R or Python, but it might not be as well set up as somebody who I might interview for like more of a intermediate or senior role, so instead of, you know, using functions and not repeating yourself, it might just be getting the job done, and if the role is really entry level, that might be okay, right, like that might just show that they have the potential to grow, so I'd say that's one, it's like kind of just assessing where their code is at, and just knowing that they can code, but it probably needs some work, right, like they have room to grow.

I'd say probably more important to me is their ability to like prioritize and take things into their own hands and take on work independently with some guidance, right, so some demonstrated experience of having done that in the past is super important, I think for any level, honestly, to give you the comfort that you know they can grow into more of that intermediate senior position.

Technical debt and tooling

Thanks, Katie, and on this topic of hiring too, if other people are like have had challenges with this or are hiring for roles, feel free to put those into the chat or jump in to the conversation too. One of the anonymous questions that came up, which was a follow-up from a previous question that I can't figure out which one it was, but it was what do you mean by incurring technical debt?

So I mean I'd say at the highest level it's just having a bunch of legacy systems that very few or no users are using, but maybe still contain important pieces of the business, so that would be like the highest level version. What I'm kind of referring to specifically is, I mean there's probably like an optimal point where you want to cut off expanding the tools, like I personally would not want to run a team from like a BI perspective where people were using Tableau, Looker, Mode, Shiny, like that's a lot of tools. Do you really need that many tools to get the job done? Oh, and by the way, Shiny is the only free one on that list, right, so now we're just spending money, but I'd say more importantly from a data science perspective is R and Python seem like close-knit communities. I do think if you're intermediate to strong in one, you can at least read and connect on the other, and if not contribute from a development perspective.

Oh, I think somebody might have gone off mute.

So, but I think what would be challenging is like if we had somebody we hired using SaaS or like MATLAB, I would have less certainty that other people could pick up their work if that person were to leave, you know, just in general, right, if their work needed a contributor to help with it on. That to me is like maybe I guess the most important aspect of not expanding too wide on the tooling perspective is like can your work be used by others, picked up, improved on while you're here or while you're gone, right, and so I think the wider your tool swath gets, the more of a risk that becomes.

can your work be used by others, picked up, improved on while you're here or while you're gone, right, and so I think the wider your tool swath gets, the more of a risk that becomes.

Building a portfolio under NDA constraints

Awesome, thank you. Yeah, Guilherme, and sorry if I pronounced your name incorrectly, I see you put a question to the chat too, if you want to jump in.

Oh, hello, I'm coming from a mechanical engineering field, so mostly we have to develop new projects or improve already existing ones, and because of this, my work comes to gather data initially and later I come to the engineering field and work, but my problem is I create these projects, I develop them, but because of non-disclosure agreements, I can't really share my work, my data analysis, so I get a lot of work, but exclusively as freelancer, because as I can't show my projects, I can't really get a job easily, so what could you guys recommend for me, like what could I do, because I have a lot of personal projects and it's not seeming like it's something that's going to improve as soon as I want it, if you could give me a response on that, it would be great.

Yeah, I was reading Zach's answer, I love that, like is it possible to simulate something akin to some of the data you would use and then feature that as a work sample? I would be curious, I do feel like project portfolios, submitting the actual work is not really something I've seen a ton in interviews, so like you need to be able to discuss the work, which might be part of your problem, I'll try to address that in a second, I've seen less often that you need to actually submit your prior work outside of like a work sample they'll give you as part of the interview process, so I would be kind of curious as to like how often you've encountered that, but yeah, I really liked Zach's answer and like simulating data, I would say if more of the challenge is what I think it might be in just simply describing the projects, I mean that's really tough, right, like even when you guys ask me about the BEAM projects, right, you want to say something meaningful, but you don't want to be too specific, so I think honestly there's probably more help than you realize in like honestly having kind of a role-play conversation with yourself on questions you're getting in an interview and practicing answers that are more generalized and abstracted, so I would take one of your specific problems and this is also a good tip for academia, right, like my academic research, nobody really cares about process outcome therapy like research, right, like that's not going to translate well, so I need to translate that to like an abstract problem that I tackled with statistics and programming, and I wonder if there's a piece of what you could do the same here, so like I don't know your specific problems, I would fumble over even trying to guess, but trying to figure out like well what am I doing in an abstract sense, am I classifying things, am I doing an unsupervised approach, am I optimizing value in some way, and trying to figure out like can you even take a stab at like verbally saying what you're doing without going into too many details and then practicing that answer, probably the most tactical help I can give there other than Zach's answer, which I really like, which would be helpful if you're, if you have to actually submit work samples, that would I think be a great play.

Oh thanks, really thanks to you and Zach, I'm starting to try to simulate where it was a bit tricky, so I was asking maybe you guys could have another idea, but really thanks, really thanks.

Yeah, and if anyone else has ideas too, feel free to jump in, I see Javier, you have your hand raised.

Thanks Rachel, yeah, I was just going to mention, I've been in a similar boat my whole career, and it's very tough because a lot of what you learn on the job, you know, you want to showcase, so how can you showcase that, and like a follow-up to Zach's comment, there's a package out there called Synthpop, and I'm sure there's a bunch of other ones, but Synthpop helps take your actual data points and sort of simulate them to make a completely synthetic data set, so like that's one way you could kind of take actual distributions from your real projects, you know, create kind of a clone of it in a way where it won't be traceable back to the source, and then create a, like a web report, you know, a R Markdown report, or a Shiny app, or something around that data. At my last job, well I'm sorry, at the job I'm currently in now, when I was interviewing for it, my team had been all around data their whole careers, but not really with programmers, so they were very familiar and accustomed with like Tableau, and Looker, and they didn't ask me to do this, but I probably spent like four hours, five hours on a weekend, and built a Shiny app that I deployed for free to shinyapps.io.

I styled the whole website, or the whole app to have styling elements consistent with this brand's, you know, this company's like logo and styling, so that they could kind of see what a really clean looking Shiny app would look like, and like I don't attribute me getting the job to that alone, but when I showed that to all of the interviewers, they were all like super impressed, like oh my gosh, we've. Did we lose Javier for a second, or is that me that we lost for a second?

Yeah, oh you're leaving us on the edge of our seat right now, but I can say I have seen that application that Javier made for that job interview, and I thought it was an awesome example, and I was just wondering if that's something maybe we could as a community help create for other people to use, where just make it like easier to show putting in certain brand colors or logos into an app.

Zach, I see you had your hand raised if you want to jump in while Javier's reconnecting. Yeah, it's fine. Yeah, so just on the portfolio front, so for me, of course, I'm looking to start my career. So I write a blog, an NFL blog, even though I'm British. That's an interesting stuff, but that's for another day. So I write all my articles and stuff, and then on my website, I write the reasons why I did it this way, why I talked about this and not this,

and all my ideas for it. So if someone came back and went why did you write it in this manner, or why did you present the graph, or why did you use a box plot instead of a bar chart or whatever, just using random examples, I could then have a list of why I did each, and so I think that really helps, especially when you're, so it's one of these things, if you keep doing it as you go along, it's much easier, because then I just go, this is my whole portfolio. If like when I start looking for jobs, I'll go, this is everything I've done. I think there's now 50 or odd so things on it now, and it just means that they can look through what I've done. There's easy links, and yeah, it just makes it a lot easier for everyone, especially I don't go to university, so I've always, so I've already got that going against me, because everyone's just looking for a degree, so that's why I came into the portfolio stuff, because I need to especially show my skills, so yeah, I'm happy to share my blog and everything, my portfolio, because I think it really helps. Also, I think Rob said about looking at LinkedIn, I think LinkedIn's also another really good place, like you can do all your features, and so people can actually see, but yeah, other people are much better than this, but this is what I've just seen and copied from people.

That's great, Zach, and I do see a few people asked in the chat if you could share that, that would be awesome. Javier, I know you got cut off a second, but we did get to hear about the shiny application that you made and using their brand colors, but would love to have you just continue the thought too. Yeah, I'm sorry, I have a parrot, and it's going crazy in the background, mainly just, you know, styling a web app with the company I was interviewing, with their entire, you know, logo set, color theme, it was all controlled through, like I used bslib, so I was able to have like a custom .css file, so sassy, you know, sass css, sass css, and like doing this kind of thing, if I was actively looking for jobs, interviewing, and interviewing, and interviewing, like the ability to iterate and change out the entire theme is so easy, I mean, it can be, once you have a template, this can be done in like 10 minutes, five minutes, so for, you know, anyone who's trying to get a job, and you're interviewing with a lot of companies, this could be a really easy way for you to have something living on, you know, your personal shinyapps.io, and whenever you interview with a new company, just like change the link out, change the color theme, and you could do this very quickly, and showcase that to, you know, your interviewers.

you know, your personal shinyapps.io, and whenever you interview with a new company, just like change the link out, change the color theme, and you could do this very quickly, and showcase that to, you know, your interviewers.

That's awesome. Thanks, Javier. Rick, I see you have your hand raised, too. Sorry, there's sirens going by. Yeah, I just wanted to mention on that point, because it's, I think, I like to think about the different ways different disciplines think about their work, and how to present it. If you've ever worked with a designer, and you get your visual identity, or you get your brand identity, or something like that, they're, they have a template, and they're just plugging them in. They're not creating all that stuff from scratch, and if you have a nicely organized project, or you have a really nice design pattern, as a data scientist, you're showing your skills as a software engineer, or really as a programmer. So, I think, like, if you, if, as data scientists, we start understanding how designers think, and how designers work with their clients, and how software engineers and programmers really work, and organize their projects, and work as a team. So, I think, like, if you're a designer, then all those things kind of come in, and strengthen those, and those, all those ideas are kind of tied in together.

Speaking at meetups and building your network

I see. Thanks, Rick. Definitely. And Katie, on this topic, as well, I know you speak at a lot of meetups, and, like, women in data science events, too, and I'd love to hear maybe more about that, if you think that could help people with building their portfolio, too. Yeah, absolutely. I mean, I, obviously, things have been a little different in the past two years, but I think some of my richest connections have come from speaking. I also think that it's just, you get, it's one of those things that's hard to force yourself to do, but you get incrementally better at it, the more you're exposed to it. So, I think it helps you just in your job, too, just present to stakeholders, present to other non-technical users, but I've found it super, super, kind of, accretive in terms of building up my own network, but not just in a kind of networking sense, actually people that you want to stay in touch with, you want to reach out to, with, you can learn from.

So, I think, like, Our Ladies, for me, is a great starting point for that. Our Ladies Columbus chapter is a little slacking, it's been a few months since a few of us had kids, we're a little slower than usual, but Our Ladies is, you know, in pretty much every city, and kind of one of the, one of the positives from our current situation is that a lot of these virtuals have, a lot of these meetups have gone virtual, so it's surprisingly easy if you don't have a chapter in your city of Our Ladies or any other meetup for you to go online and find a meetup you can join, and it's, it's been cool to see people from all over the country come to our Our Ladies meetup, and we've had some guest speakers from Canada that we would have never, ever been able to expose our audience to, so I think it's super valuable, and it should be fun, if it's not fun, right, like, you know, then it shouldn't feel like a pain to do it, right, so find the right meetup for you.

Definitely, and I can share in the chat, someone made a great application where you can see, like, all the meetups around the world, I'll get that in just a second, but Libby, I appreciate your comment in the chat, too, about, like, convincing ourselves to overcome imposter syndrome to speak at meetups is, it's kind of hard sometimes, and I was curious, Katie, if you have any tips for that.

Yeah, I would say start with a comfortable one, so I know people are pretty hesitant to speak at our Our Ladies meetup, and so we try to have people who were hesitant be then the advocates for people who are hesitant once they've spoken, so if we have somebody who was, like, a little scared to speak, and then they did, they had a great experience, we help them kind of have people shepherd people into speaking, but it, if you're interested, we are always looking for guest speakers at Our Ladies Columbus, so definitely feel free to hit me up, but I would say start with a small group, right, a small comfortable group where you know it's kind of low pressure, right, so, like, Our Ladies, at least Columbus is a good example of that. Sometimes the presentations are just walking through code, right, it doesn't need to be super formal, and it's ideal for us to get somebody who's learning something, because oftentimes when we have an expert, they just breeze through a bunch of really complicated concepts, and nobody can really follow them, but start simple, keep it frequent, so Toastmaster is obviously non-technical, but another way to get, you know, exposure to speaking, but I really think all speaker coaching is that exposure effect, in my opinion, I think it's just doing it, getting feedback, and you know, realizing that nothing terrible happens even if you flub.

Definitely. Thanks, Katie. Jaiwan, I appreciate that tip in the chat as well, if you'd want to jump in and say that out loud. Yeah, just wanted to recommend the five-minute lightning talks at various different user groups, and I think the last one was that even like a per package that everyone really knows about, and put tips on it, so yeah, I think those are really good opportunities to speak up and practice your presentation skills. Yeah, I think that's a really good opportunity to practice your presentation skills. Definitely. Thank you. That's helped a lot for the Boston Usar group too, and even as like a community organizer, if you can't find people to speak sometimes, like doing lightning talks is really helpful.

Remote vs. on-site work

I wanted to make sure I got to some all of the anonymous questions that came in too, and I just noticed there was one, Katie, that was how do you feel about 100% remote versus on-site?

Good question. So our team is a hybrid right now, and I think that's kind of the trickiest format, and I feel like we've done pretty well in terms of making that work. I guess is the question more about like my preference, or just managing a team that is one or the other? I'm not sure, but I think it might be your preference, and then I would love to hear how you do that with the mix of remote. Yeah, I personally love a hybrid remote, like a hybrid approach if there are other, if there is an office to go to, if that makes sense. So for me, just the role I'm in, I feel like if there are stakeholders meeting in person, I do feel like it can be a little challenging if you're the only remote person, or if they're frequently there. Now full remote, I also think I could totally adapt to if a company was full remote, right? But I do think kind of the hybrid can be a bit tricky depending on your role. We have several individual contributor technical roles, and a lot of them are completely remote, rarely come to the office, and they thrive, they do great, their work is, you know, just as impactful as anybody else's. I think it depends on the company, but if you're in a role where your peers or people you are serving are in person on some cadence, it can be difficult if you're never in person. That's just my take. In terms of how we do hybrid, though, I think the key for me is, like, we should never have a non-remote friendly event. That's just not fair, right? So, like, we should never have, like, happy hours that are in person where other people aren't included, like, team fun, happy hours, those are virtual, so.

Thanks, Katie. And Brick, I see you had your hand raised for a question, too. Sorry, that was from before. I didn't lower it. Oh, okay. No worries.

Graduate degrees vs. modern tool stacks for hiring

Just wanted to check if there are any questions I missed from the chat. Feel free to jump in and unmute yourself as well. I just had a question previously on the hiring aspect of data for the data science role, because I've been thinking about it myself, what would be the best next step, getting a master's or a graduate certificate in statistics, versus getting more involved into modern stacks, or modern tool stacks, and maybe big data technologies? And I was just curious, when we were talking about hiring process, what would be your, I guess, preference? And I understand it's under use case basis, but I'd be curious to hear your thinking about that.

Yeah, good. I mean, my answer, of course, would be it depends. I would say there are definitely a lot of degrees popping up from, like, master's of analytics programs, master's of data science. This is probably just a personal preference thing, but I would probably maybe favor somebody with a master's of applied stats over those programs right now, just based on where our team's at. So I feel like those programs are a little broader versus deep. And we have kind of a good setup on the team of people with that broad exposure to a variety of technologies. I think it's probably, from my perspective, would be a miss to hire somebody who has dabbled in, like, big data platforms versus just going after a proper data or analytics engineer, right? Like, I'd rather hire somebody that's deep versus somebody who's kind of dabbled in a lot. And I'd say my same principle probably generally applies for statistics.

I think it's very easy to have, like, a cursory understanding, but I think you have a different level of training and rigor if you've gone through more of, like, a statistics-focused program than kind of the general level. But that being said, there are, of course, people who are fantastic in both of those aspects deep on their own, right, because they've taught themselves that, right? Like, they're just fully capable of learning on their own, and I never rule that out either. And I guess as a follow-up question or the context that I wanted to add to that, I've heard one of the opinions of a data mentor that I respect was that these days the depth, the in-depth knowledge of statistics may not necessarily be relevant, and the superficial may do, especially that now you can run, like, a thousand classification algorithms on your laptop that would give you the response without you necessarily knowing all the ins, like, without you necessarily having an active role into choosing which model would perform best, looking at the, all of the, you know, performance indicators when the program can do it for you.

I'm like, well, that's a good point. I honestly, this is a dilemma I found myself in when hiring. I was like, I need a PhD in stats, and I'm like, do I, do I really, right? Like, and I didn't, I don't, and we didn't get a PhD in stats. We have two extremely capable, very strong data scientists. Both of, both of them have masters in kind of engineering-like fields. One's bioinformatics, which obviously has a heavy statistics training, so it kind of depends on the company. For me, at our company, you need to know enough to be a informed consumer of those packages, right? To not, I need to trust you enough that you know enough that you're not going to put something out there that is wrong or that I have to follow every single detail of what you're doing, but that being said, there are certainly R&D shops, companies where the core of their business prop is telematics data, is, you know, very niche custom algorithms, and that, I think, is where you do need that in-depth training. So, of course, kind of depends, but I think what probably is not best is if you've had like one or two undergrad statistics courses, and you're using like this Cadillac package that allows you to do anything you want, and you really have no idea what you're doing, so that's basically what we try to avoid in a nutshell.

I think what probably is not best is if you've had like one or two undergrad statistics courses, and you're using like this Cadillac package that allows you to do anything you want, and you really have no idea what you're doing, so that's basically what we try to avoid in a nutshell.

Yeah, thank you. Thank you, Katie. I realize once again an hour went by very quickly. I just wanted to check with you if there are certain resources or like books or podcasts that you'd recommend to us all here that have been especially helpful for you. Oh man, there was a series of videos on our bloggers called 14 Hours of Machine Learning, and it was the slides and lectures from Hastie and Tibshirani from their book, like Intro to ML. I'm going to try to, I'll send it to you afterwards. They've taken down the links to all the free videos

somewhere. I can't find it anymore, but I will look for that. I think that is like the best intro to classic stats problems resampling. I'm a big fan of all of the like RStudio articles out there, so I think like Tidy Models, for example, that whole suite on their website really allows you to get up and running super quickly, and then obviously Meetups, and the proponent of those two I think you can learn a lot, so hopefully that was a quick but informative answer, but I just love that Hastie and Tibshirani resource for like the fundamentals. That's really helpful. I put one into the chat, but I just googled really quickly and didn't read it, so that might not be it, but that is, but the links, it was, but the links don't work, so track it down. Yeah, well, we will find it somewhere, and I'll share it in the LinkedIn group as well, so you also have the data science hangout community LinkedIn group if you want to join that. I try every week to put in a few snippets from the conversation, but thank you so much, Katie, for jumping on and sharing your insights with us and answering all these questions.

If people do have follow-up or want to connect with you, is Twitter the best way or LinkedIn? What works best for you? Probably LinkedIn is probably the best source there, yeah, on Twitter, but usually when I go to a conference, I'm not the best Twitter steward, so. Yep, okay, same with me. I'm learning to be better at Twitter. Yeah, I don't have the skills, but I'm going to share the data science hangout LinkedIn community there as well in the chat. Thank you all so much for joining. Have a great rest of the day, guys.

Bye.