Grammarly, the Chrome extension that helps people write right, bases its spelling, grammar, and word use suggestions on intense analysis of vast databases of the written word. One of the people analysing that data is Dr. Courtney Napoles, a computer scientist with a background in book editing. We talked to her about her work and expertise, her morning routine as a single mother, and what she loves about the classic Emacs text editor.
Location: New York, NY
Current gig: Language Data Manager at Grammarly
Current computer: MacBook Pro 13"
Current mobile device: iPhone 8
One word that best describes how you work: Multitasking
First of all, tell us a little about your background and how you got to where you are today.
My professional background is a bit unique because several years ago I decided to make a big jump and switch careers entirely. After graduating from Princeton University with a BA in psychology, I worked in the trade publishing industry, editing “helpful” non-fiction books. As surprising as it may sound, this inspired me to pursue a doctorate in computer science at Johns Hopkins University.
Now I specialise in natural language processing (NLP) — a collection of techniques to help machines understand human language. NLP excites me because my work helps millions of people around the world every day.
What attracted me most to Grammarly was its mission-oriented drive as well as the strong communication focus—the company is dedicated to helping improve people’s lives by improving their communication, which is incredibly motivating.
When I first started working here, I was hired as a research scientist. Now I’m the language data manager and am responsible for building and leading a team of analytical linguists to create and evaluate linguistic resources for NLP, deep learning, and machine learning models.
Take us through a recent workday.
As a single mum, my day starts by getting my son ready for school. Before we rush into our morning routine, however, we take a couple of moments to cuddle and chat in bed, which is a calming way to begin our day together. I savour these moments because I know these memories are fleeting!
Mornings are especially hectic, and heavy multitasking is required to get everything done. While my son watches TV, I make him breakfast and myself a coffee before checking email or Slack. Then I prepare him an after-school snack, walk the dog, and get us ready to head out the door.
By the time I get to my office, I’m ready to step into my second role of the day. I typically begin by catching up with my team while sipping on an Americano. After lunch, I tend to stand, which allows me to get the blood flowing and dive deep into the details of a project.
Meetings take up the majority of my day, but I spend as much time as possible training new folks on how to evaluate data within our algorithms, conducting high-level planning for the next research project, or onboarding new team members.
What apps, gadgets, or tools can’t you live without?
I would be almost paralysed at work without Emacs, which is a powerful text editor that is popular among programming and software engineers for writing code. I use it in “org-mode” to develop to-do lists, handle shifting priorities, organise my notes, and tag necessary people.
I’ve only been using it for a year, but ever since I started I can’t take notes anywhere else. And not just at work but in my personal life too: shopping lists, packing lists, tracking my son’s summer camp schedule, and more. It’s so versatile and powerful — you can use it for anything!
What’s your workspace setup like?
I work out of Grammarly’s New York City office, which is currently in a great WeWork space with plenty of natural light, open seating, and private conference rooms. My seat is next to a window, and if the NYU Hospital weren’t obscuring my view, I would be able to see the East River!
My desk is a sit/stand desk (which I take advantage of for my post-lunch standing time). My personal workspace includes lots of natural light and plants. I also have an ergonomic keyboard, which is incredibly helpful for programming — but also prevents me from typing loudly, as I’ve been told that I type with purpose! There’s also an external monitor, sticky notes, and a legal pad with plenty of pens.
My favourite type of pens are Paper Mate Flair pens. Although I don’t frequently use pen and paper to complete my job, I love having these pens by my desk. I find that writing with the fine felt tip is a great mental break.
What’s your favourite shortcut or hack?
In general, if there’s a way of maybe optimising my workspace or desktop, I’m trying it. But I have two shortcuts that I’m very conscious about using. The first is that I try to avoid using my mouse as much as possible. I find that I’m able to work more quickly with keyboard mapping, which is when you customise your keys and match them with specific commands and actions. So I’ve created shortcuts for everything that I have to do on a day-to-day basis and cut out a lot of extra steps.
I’m also devoted to the app SizeUp, which helps me quickly resize and position my open windows in a way that optimises the size of my screen. As a result, I save time and frustration by avoiding switching from tab to tab. This is especially significant because in my role I have to analyse and compare large sets of data. The tool is perfect for helping me visualise sets side by side.
Who are the people who help you get things done, and how do you rely on them?
One person who has been a major part of my career at Grammarly is a colleague in our San Francisco office. We don’t work together directly, but we briefly connected a year and a half ago and bonded over being working mothers in tech. Since we met, we’ve been video conferencing every other week and often talk about work-related opportunities or challenges. I’ve found that having this support from a like-minded person with similar goals has helped me tremendously in being able to evaluate job situations because she has a completely different perspective.
Helping others is also another common bond that we both share, so there’s an added layer of empathy to our conversations that is sometimes lost in peer support groups. After we talk, I often feel more confident and prepared to control any hyper-focused thoughts that are giving me trouble or are preventing me from getting my job done.
How do you recharge or take a break?
One way I recharge during a busy day is to get outside and walk to a favourite park that’s close to the office. I don’t do this as nearly as much as I should, but every time I invest time in even just one lap around the park, I feel instantly revitalized.
What are you currently reading, or what do you recommend?
I’ve experienced a lot of changes in my role this year. I’ve been tasked with leading and building out a brand-new team. I’ve also become the site lead of our NYC office, which has brought on new challenges for my management and leadership skills — in the best way possible. To stay ahead of these situations, I’ve been reading The Making of a Manager by Julie Zhuo, and so far I’ve been able to apply it to multiple different situations, such as recently hiring and on-boarding two employees.
What’s the best advice you’ve ever received?
If you’re unsatisfied with a certain experience or situation, always look to yourself first before turning to others. This allows you to interpret the problem from your perspective first, and you’re more than likely able to either solve the problem yourself or at least identify what needs to change to fix whatever’s bothering you.
What’s a problem you’re still trying to solve?
Grammarly built its reputation on being a grammar checker that helped you fix mistakes in your writing, but we offer much more extensive assistance than that and are always building. We want to keep expanding the ways we support people in their communication beyond helping them build confidence with language rules. Where we’re continuously pushing our research is toward assisting our users in writing more clearly and effectively — ultimately striving to help people feel heard and understood.
Working on an engineering product with such human goals is definitely not an easy challenge. Grammarly is powered by AI — a combination of rule-based, deep learning and machine learning approaches. Some of these methods are very data-hungry, meaning they rely on learnings from thousands or even millions of annotated data. I lead a team at Grammarly dedicated to creating these large-scale datasets for training our AI models. The eternal problem that we face is: How do you handle the trade-off between quality and available resources?
We are very concerned with having high-quality data — especially data that doesn’t introduce or perpetuate bias in language. I work closely with our research engineers to make sure that the data we produce allows them to create the models we expect — all while working in an Agile framework.
In addition to these responsibilities, I’m building a global team that has already doubled in size since I joined, so I need to focus every day in maintaining tight collaboration and relationships across time zones and continents while also being on the search for a wide range of skills to add to our growing organisation (linguistic expertise, project management, coding skills, and knowledge of NLP).