Andrada is a passionate data scientist who engages in voluntary research projects around the topic of language on the internet. What she knows for sure: all opportunities deriving from data science leave no space for redundant inequalities and biases.
Women are equally capable and driven to succeed in sciences. A deep societal change is needed to bring awareness to gender biases and convert more people into allies. Until that happens, the least we can do is support each other. So I want young women to know that there are women out there just like them, with the same background and likely the same doubts who fought hard and they will need to fight themselves.
I’ve always imagined my life in a lab, conducting experiments with tubes of different shapes and sizes and nodding approvingly (or not). Data science brings me the same joy (and challenge) of designing experiments, and holding my breath to see if my assumptions were right or not. The field implies a steep learning curve and requires a diverse skill set. It’s exciting to stay on top of things in a fast-changing field and I find it particularly appealing to learn numerous new techniques and methods.
On a more realistic note, data science is often over-hyped, with companies expecting it to be “a magic sauce” that will propel them forward. While designing experiments is fun and machine learning has the potential to solve specific problems, most industries need data engineering and deterministic data pipelines first – to make sense of the chaos of collected data that “we might need later”. Going through this process is equally satisfying and often calls for creative problem solving.
I am very privileged to work in a company like Futurice, which genuinely cares about gender equality. I especially appreciate our company’s continuous involvement in D&I initiatives and its strive to be better. In general, data science communities that I am part of in Berlin are very open-minded and attentive to correcting gender biases and dynamics.
There is a very subtle bias that occurs from time to time. The most prevailing issue, that I often find myself fighting, is being invisible. It’s a double fight: first being acknowledged, second proving that indeed, you know what you’re talking about. What is truly disheartening about this, is that when you call this out, the response will be that you are not assertive/confident/loud enough.
Women are equally capable and driven to succeed in sciences. A deep societal change is needed to bring awareness to gender biases and convert more people into allies. Until that happens, the least we can do is support each other. So I want young women to know that there are women out there just like them, with the same background and likely the same doubts who fought hard and they will need to fight themselves. Having role models is very important and provides an extra boost of motivation. From my personal experience, it’s very important to find a mentor as early as possible, who can support your professional journey and personal growth. I have a mentor at Futurice and one outside of Futurice. Having somebody to spar with is incredibly valuable, especially when they challenge your own assumptions and push you to become a better version of yourself.
I’ve always been interested in languages, so I specialized in NLP (natural language processing). Language defines our experience of the internet. To make the internet more inclusive, I want to contribute to the development of technology that is fair and accurate for any language. Last year, I joined a project called Opt Out and have been organizing their data team since. Opt out is a browser extension hiding misogynistic tweets from social media feeds through machine learning. The team is lovely and diverse and has considerably grown over the last few months as we’ve been collaborating with experts from linguistics, social sciences, content creation and engineering.
Besides, I worked on another research project on hate detection in different languages last year. I realized that there is no research in automatic hate speech detection in my native language Romanian at all, so I created the first labelled dataset. I presented the results at the PyCon & PyData Berlin 2019 Conference and several meetups. Finally, I’ve been working on training a Romanian language model boosting NLP tools together with a startup in Berlin.
Giving a talk about a personal research project at a big conference was definitely my proudest moment from last year. It allowed me to collaborate with incredible people and connected me with peers with similar interests, dedicated to fighting against social inequalities. Technology alone is not sufficient for creating lasting change and positive impact, but instead a more holistic approach is needed, which includes humans with various backgrounds and life experiences.
Interested in reading more stories about us and our people? At Futurice, we celebrate diversity and cherish everyone's unique journey. Check out our Welcome Home page and get inspired by more journeys shaping our culture. If you would like to read more stories and get to know our people, our sites and the community better, check out the global version of our FutuStories Booklet.
