Travis Gerke | What they forgot to teach you about industry transitions from academia (WTF AITA)
This talk presents a gathering of resources from the RStudio community for industry job-seekers who are transitioning from academia. Examples include packages for resume-building and cover letter writing, quick starts for personal websites and GitHub profiles, and helpful R networks. A new website named after the talk (wtf-aita.com) will showcase and link to the discussed resources, and is open to further community contributions. Session: What they forgot to teach you about your career
image: thumbnail.jpg
Transcript#
This transcript was generated automatically and may contain errors.
My name is Travis Gerke, and I am the Director of Data Science at PCCTC, a clinical research organization facilitating clinical trials in the cancer space. I used to be an academic, and when I was transitioning to industry roles, I found challenges in refactoring my resume and professional portfolio to align with industry standards. I hope to share some resources that you'll find useful in this talk, what they forgot to teach you about industry transitions from academia, or, more simply, WTF AITA.
If you're not familiar with these two acronyms, we'll consider each in turn. First, if in the context of your job you've ever asked, what the fuck, I'd probably respond with, great question, particularly if you're an academic.
This plot shows one aspect of WTF in academia, the increasing number of PhDs awarded versus the much smaller number of faculty jobs available. The gap between these two diverging numbers has helpfully been labeled WTF. There are lots of other reasons one might ask WTF in academia, but this is not really a talk about that. I would instead refer you to the so-called quitlet, where people have discussed their thoughts and experiences on leaving academia in a variety of formats, such as blogs and social media posts.
Indeed, academia can be a highly rewarding career if it's what you want, and this talk is not at all focused on criticizing that career choice. All the same, many academics are currently industry curious, and I hope that many who are not in academia will still find useful content in this talk. Academics certainly don't hold a monopoly on asking WTF.
The AITA problem for academics
The second acronym, AITA, stands for Am I the Asshole? This happens a lot when academics first apply to industry roles. As a specific example, I was asking this question when my first job applications to industry roles received no response. I thought, am I not qualified? Are they the asshole? Am I the asshole?
In this talk, we'll focus on three common ways that academics can unintentionally be the asshole. First, resumes that are way too long are common. Second, cover letters should not be a thesis on the direction of your research, rather a clear and focused statement of your value that is aesthetically streamlined. And last, an untidy coding portfolio can make it very difficult to see how the first two items come together to make you amazing.
In short, avoiding assholery really means not asking others to work hard to discover how awesome you really are. As a current reviewer of many application packages, I see some truly excellent candidates who could shine that much brighter with some finesse. Fortunately, the R community has created resources which make each of the three problems easily solvable.
Resumes
Here, quite embarrassingly, is the 22-page resume that I sent, as it is, to the first industry roles to which I applied. This beast served me well in academia, where CVs are required to be formatted like this. In industry, however, yes, I was the asshole.
Here on the other hand, is a stellar resume that I received about a year ago from someone a lot of you probably recognize as this conf, Shannon Pileggi. This elegant two-pager gives me an easily accessible view of Shannon's skills, experience, and clear links to her portfolio, a topic that we'll discuss later. I'm lucky enough to work with Shannon now, and she gave me permission to use this as an example.
If you want your resume to look like Shannon's, fortunately, there's an R package for that. It's called Data-Driven CV. I won't repeat any of the how-to content that the package author Nick Strayer has carefully assembled. If you go to the package vignette, you should be up and running in under an hour.
Aside from making a resume that looks really slick, there's a big bonus with using a package like this that I want to mention. When hiring managers like myself fit a stack of 50 to 100 resumes, I can recognize right away those applicants that use R to create their materials. This practice sends a signal that you're excited about integrating R-based workflows into all of your activities, not just those that involve statistics or data analysis. And it's a sure thing that I'll read those resumes carefully.
When hiring managers like myself fit a stack of 50 to 100 resumes, I can recognize right away those applicants that use R to create their materials. This practice sends a signal that you're excited about integrating R-based workflows into all of your activities, not just those that involve statistics or data analysis. And it's a sure thing that I'll read those resumes carefully.
Something else that the Data-Driven CV package does that's nice is that it makes it easy to include or omit information for long or short versions of your resume. For instance, if you still need a long academic variation for some purposes, you can render with all the publications and fluff into dozens of pages. And then if you want the industry-ready two-pager, a quick filter within Data-Driven CV will get you there. This eases some of the pain I experience in moving the currency academia pages and pages of publications to a single bullet point in my resume.
Cover letters and cover accessories
Much like my horrific Academic CV, I was also sending out this dense research statement as a generic cover letter. Here instead is a template which is easy on the eyes through generous use of whitespace, is succinct at one page, and even makes use of bullet points to highlight key ideas. Recall the signal sending I mentioned about resumes clearly built with R. The same is true here. I right away recognize this is R markdown or LaTeX rendered due to general font and alignment structure. A dead giveaway is the appearance of awesome font icons in the top right. That typically doesn't happen by accident in a Word doc.
If you want to render letters like this, here's the steve Templates R package from Steven V. Miller. Making one of these is as easy as installing the package and creating a new markdown file from template in the RStudio menu.
But cover letters aren't the only thing you're allowed to send as part of your application package. Let's have a look at something really cool in the category that I'll call cover accessories. In a recent RStudio Data Science Hangout, which is an awesome community resource that I'll describe later, I met Javier Oraca. He described his recent application to BloomReach in which he went way above and beyond and made a BloomReach Shiny app with some simulated data that is themed with their org's color scheme and logo. The BloomReach team was, of course, totally blown away and Javier landed the job. His technique of demonstrating what's possible with Shiny as a business intelligence tool along with his mastery of developing in it was spot on.
If you're interested in the secret sauce here, it's two packages used inside of Shiny, bslib and thematic. I'll mention that if you're not a Shiny developer and the idea of integrating these components into Shiny isn't something you want to tackle right now, don't forget that you can use these same principles, namely organizational theming and showcasing your R skills within a standard R markdown document that is rendered to HTML or PDF.
Portfolios
Here's a common look for an academic coding portfolio and yes, this page is intentionally left blank. Academia doesn't incentivize maintaining this sort of thing, but it's hugely helpful in showcasing your technical skills for industry employers. I'm going to show three levels of portfolios with increasing complexity. Only one is really necessary, but it's certainly possible to have all three.
First, not only is github.com the standard place to deposit your projects and code for public viewing, the site also allows you to put together a personal profile on your landing page. Here's a great example from Monica Powell and her link here shows you how to create this using the GitHub interface. Here though, we're going to see how you can do it in RStudio. Conveniently, it's just three lines of code.
The steps are this. Create an RStudio project with the same name as your GitHub user ID. Connect it to GitHub and create an R markdown document that renders to GitHub compliant markdown. Again, these three lines of use this code achieve these three tasks directly. Mash the knit button in RStudio, push to GitHub, and you're done. Any modifications you'd like to make are just done with R markdown syntax that you probably already know. Remembered R plots and all the bells and whistles work here.
As a super basic example, here's what happened when I simply rendered the use this RMD default document through this method and pushed it to my GitHub repository named my user ID tgirky. Cool, right?
While we're on the topic of GitHub, I want to emphasize that you should use and organize your pinned repositories that appear at the top of your page. If you've done a solid job on the resume and cover materials that we discussed, it's a virtual lock that your GitHub profile will be looked at. Put good code that you're proud of there. If it doesn't yet have good documentation or comments, add them now. Yes, the very careful reader might be able to tell that you added them after the fact, but I claim that this is not a big deal. Adding them will show that you understand the importance of well-documented code and that you can produce it.
Taking this all up a notch, here's our friend Hadley's landing page at his website. Looks awesome. The great news is that you, too, can be Hadley. This is made with Sean Karras's excellent postcards package. Hadley used the Trestles template, but there are other great-looking options here if you want to branch out a bit. Getting started is made very easy with the help of the package vignette, which is linked here.
Now here in expert mode, you can develop a full-blown website with a blog using RStudio, like Allison Hill has been doing elegantly for years now. Her most recent iteration uses Hugo Apero through BlogDown, which she will carefully walk you through in detail in the linked recording of her RLadies talk from last year. BlogDown with Hugo is one way to do it, but other popular options include Distil and Quarto. Danielle Navarro has superb examples of both in the form of a Quarto blog that was migrated from Distil. She describes each of those formats and how you develop in each at her blog post, which is linked here.
When it comes time to publish your content on the web, don't forget that you can use free services from GitHub Pages and Netlify. There's no need to purchase a dedicated domain name if you're not ready for that yet.
The R community
Overlaying all the topics we've discussed is the incredible R community. If you ever need support, want to discover who is hiring, or generally get connected with amazing people, here are some directions I would suggest, all linked from this page.
In order, Rachel Dempsey runs a virtual weekly RStudio Data Science Hangout with data science leaders and community members. This informal get-together is an excellent way to stay in touch and find out what's new in the R and data science world. Tidy Tuesday is both a podcast run by John Harmon and a highly engaged weekly community activity run by Thomas Mock. It's a great way to hone your skills, show off your data science wizardry, and learn from others. Tidy Tuesday is just one component of a larger, very active R for Data Science community.
This community runs a Slack server where you can ask and answer technical questions or just chat about data science in general. One very helpful channel there is a jobs hiring board where you'll see many R-specific jobs posted.
R user groups, or so-called RUGs, can be found nearly anywhere in the world these days. If there isn't one close to you, consider starting one. The R Consortium sponsors many of these and will help you get started. Back when I was in Tampa, I started one of these which met monthly, and I met friends which I'm very close with to this day. Many job-related connections tend to happen at these events as well.
In the same vein, R-Ladies Global is a worldwide organization that promotes diversity in the R community. Much like the RUG meetups, you can probably find one of these running in your area and you should definitely check it out. Lastly, the RStudio community has forums and job boards that range from the technical to community building and is a great place to hang out.
WTF AITA website and closing
The final resource I'm going to leave you with is a compilation of all the resources we've discussed and more. Wtf.aita.com is a place where you can find these slides and links to the tools therein. But this talk certainly only represents a slice of a much larger set of guidance and resources around industry job-seeking. Many of you in the RStudio community have already traveled this road, and if there are things you'd like to see represented on this site, please do feel free to submit an issue or a pull request at the GitHub repository. It's my intent that the site organically grow as a community-owned utility over time.
It's my intent that the site organically grow as a community-owned utility over time.
With that crowdsourcing request in mind, I'll now hand the discussion off to the excellent RStudio community. Thanks so much for having me at conf, and I look forward to seeing you in the Discord hallways and answering any questions that you may have in this particular session.
