We are incredibly excited to go live with our conversation with the amazing Stacey Ronaghan.
This conversation is a must-read for all everyone looking to get into this field and stay engaged and active. Stacey gives you actionable advice with candor based on very relatable personal experiences that will encourage you to go forward.
SDD: Stacey, can you tell us a bit about your journey in data science? How did you get interested in this field at first and what steps did you take to pursue it?
Stacey: I’ve always enjoyed mathematics so decided to study it at university. I considered it to be a good platform for whatever career I decided to pursue. However, when I graduated, I didn’t actually know what to do with the degree. The only companies that had visited our campus for recruitment, or at least I came across, were accountancy and actuary science firms, neither of which enticed me.
My first graduate role was in market research, writing reports about sports and medical electronics. I found this role interesting as I’d interview people, write reports, and present my findings, substantially growing my soft skills. However, I itched to do something that involved more analysis. This led me to my next job as a technology research analyst for a broadband company. It involved business intelligence and competitor analysis. The role required that I refresh my R skills, which I studied briefly during my semester abroad. When I was asked to write an internal whitepaper on Big Data, I decided this topic was something I wanted to pursue. IBM was referenced a lot, so I decided to apply there, knowing they had a program for new graduates that didn’t necessarily require a computer science background.
I was conscious that I’d “taken a step back” to have this career change, particularly as I was a few years older than most of the others on the program. However, my prior work experience enabled me to quickly provide benefit to the company whilst I received support, and hands-on experience, to learn to code (this new role certainly highlighted how limited my R skills had been). My first position in the company was to look at offloading current product capabilities to GPUs. As a result, my first programming language in the company was actually CUDA C – quite the introduction! I don’t recall how, but I discovered that machine learning algorithms could be parallelized, and I researched this to provide the product with new functionality. This initiative was encouraged, and I started learning machine learning algorithms and converting C++ implementations into CUDA C, testing performance improvements. I was now introduced to the term Data Science and once again felt the need to follow this path.
I moved to another team in IBM where we provided customer solutions using big data technologies. In order to skill up on this, I took many Coursera courses, starting with Andrew Ng’s Machine Learning course in January 2014. To this day I am so grateful for online courses as I find myself much more engaged with videos and visuals and the ability to learn at my own pace. I expect I will always have a preference for videos over whitepapers. In this new team, I became a consultant and data scientist, working with clients to showcase the capability of open-source and IBM proprietary big data tools.
Since then, I moved from the UK to San Francisco and now Los Angeles. I’ve worked with many data science tools and frameworks (e.g. Spark, TensorFlow, and Watson Studio), as well as many programming languages (Scala, Python, R, and SQL). What keeps me interested in the role is the ability to continuously learn, not only gaining familiarity with new tools, languages, and algorithms but also business domains. As a consultant, I get invited to learn about my clients’ domain whilst sharing my knowledge and expertise.
SDD: There are many women in our community both freshers and women who are looking for career switches who are keen to enter the world of data science. What is your advice for them on how to get into this field?
Stacey: There are so many great university programs, online courses, and boot camps that provide the foundation of data science knowledge. Those with a capstone project or internship opportunities are particularly great, as there is nothing like working with real-life (usually pretty messy) data.
If I were to repeat the process again, I would certainly spend more time networking. It is actually something I’m trying to work on now. There are so amazing groups of fabulous technical people, many dedicated to women and gender minorities, such as R-Ladies, PyLadies, Girls in Tech, WiMLDS, and AI Inclusive. Also, meetups and conferences are a great way to learn about real-world data science use cases and companies. The big tech companies are well known but there are so many smaller companies working with data science and looking for talent. Meetup events can help you find companies local to you that would benefit from these great skills you’re learning!
I don’t feel I can answer this question without touching on the dreaded subject of interviewing. I experienced my first notorious “data science interview” a few years ago in San Francisco. Although I read about what to expect, I wasn’t prepared. I enjoyed the take-home challenge but when I was at the on-site, I got flustered and didn’t build anything particularly useful and couldn’t answer questions on how a particular algorithm worked. I was too embarrassed to talk to people about my experience as I had already internalized it as proving I wasn’t a real data scientist (hello, impostor syndrome!) and decided I would never apply to another job again. Now I know this is part of the process, like anything you can’t expect to be perfect the first time. Since then, I have interviewed at a few companies and had positive experiences resulting in a mixture of offers and rejections. Interviewing is hard but you will get better with practice and learn to identify what you want from your future role.
This leads me to my final piece of advice: be kind to yourself. I always thought I’d need to be the most deeply technical person, reading whitepapers at breakfast, to do well in data science. I still worry to this day that everyone knows more than me and is solving far more advanced data science projects. However, this is due to insecurities rather than reality. Sure, there are going to be people that are smarter than me or those who are solving more complex problems, but that doesn’t belittle the work I’m doing. I’m grateful for my ability to build relationships with my clients, by being inquisitive and not shying away from asking questions to ensure I really understand their requirements, as well as my persistence to learn the required skills to solve their problems. Data science is a broad field, you can’t know everything, so work to your strengths and the things you enjoy. If you’re transitioning into this field, it is wonderful you have found a domain that excites you and you’re already gathering the skills you will need.
SDD: Most women in our community enroll in online courses to learn data science and complete them. However, once the course is over they are unsure about how to apply those skills. Apart from finishing these online courses, what else should they be doing so that they can develop skills that can be readily applied to projects?
Stacey: I’ve many colleagues and friends that adore Kaggle. If that’s something you wish to use, I recommend creating a team, either with classmates, colleagues, or virtually, to keep you accountable (making sure you put the time in) and to help each other (providing support and advice if someone gets stuck). Platforms like this can give you experience with different types of data science problems you might not yet have experience with. If you want to try building a computer vision solution or work with time-series data, you can find appropriate data sets with clear objectives.
If you’re keen on having a side project, you might be able to find these through a meetup of like-minded individuals. I have an interest in using my data science skills for social good so in San Francisco, I volunteered with Code for San Francisco (part of Code for America) and DataKind. You might use Hackathons, often advertised within meetup groups, as a way to use your skills, learn others, and meet new people. Collaboration is incredibly important in data science; it is a team sport and you’ll need to work with a range of different people to be successful, so any chance to work in groups, I’d highly recommend it.
If wishing to make a career change, I’d recommend looking for internal opportunities to exercise your new skills; try to find teams that required some data analytics. You may find that you’re not immediately doing machine learning but that doesn’t mean you’re not doing data science. The quickest way to help other departments is likely to start by exploring their data in ways they’ve not had the opportunity to do yet. If you can ensure you are getting the right questions answered, this can be incredibly impactful. Any experience working with real business stakeholders and industry (again messy) data will be valuable.
SDD: Are there any recent advancements in Machine Learning that excited you? Any papers or projects that particularly caught your interest?
Stacey: I’m really excited about work on ethical AI. This is so important to ensure that what we’re building isn’t biased, causing harm or further reinforcing stereotypes. Not only are people speaking up about companies that have released harmful products (e.g. sexist credit limits, racist recidivism predictions, and recommendation engines promoting fake news) but they are working on ways to recognize and mitigate these issues with education and awareness, design practices and toolkits. Similarly, explainability is another topic that interests me. In order to trust AI, we need to be able to understand why it makes particular decisions and ensure that it aligns with our values.
SDD: What inspires you every day?
Stacey: People. I’m a very social person, so I love to meet new people. Sometimes I need to psych myself up to go to an event but when I do, I’m always energized and excited. I’m actively trying to expand my network socially (with my recent move to LA) and professionally (as highlighted above, this is important and inspiring) and by doing so I’m rewarded with stories about work, hobbies, challenges, and successes. For anyone else interested in building their network, they are welcome to connect with me on LinkedIn mentioning She Drives Data in the invite.
Join our She Drives Data community on SHEROES (You can also download the app on iOS and Android) to connect with thousands of women data enthusiasts across the world, hear from data science experts and get updated on the latest tech news every day ❤️️