At Echobox, we produce complex AI and content automation. This year alone, we’ve achieved some incredible milestones, from celebrating Echobox Email’s one-year anniversary to automating more than 3 million work hours (a staggering 72% increase over the amount of work automated in 2020).
We also pride ourselves on finding amazing talent from around the world. So we decided to introduce you to some of the people that make Echobox technology possible in the first place. This week, we’re talking to two of the data science masterminds behind the Echobox Email algorithms: Anna Hlédiková and Ilaria Gallo. We’ll discuss their role at Echobox, the good side and the less glamorous side of data science, their background, and their favorite national fare.
Ilaria grew up in Padova, one of Italy’s most charming and oldest cities. She remained in her hometown for university, studying in – you guessed it – one of the world’s oldest educational institutions. After a brief stint at Sony, in Germany, she joined Echobox in 2020 at our London headquarters.
Anna hails from the beautiful city of Prague, Czech Republic. After moving to London for her university studies, she joined Echobox in 2021. She’s also published award-winning research (keep on reading to find out what!).
Without further ado, let’s join our data scientists for a fireside chat.
Echobox: At Echobox, we often take a unique approach in many aspects of our business. What would you say sets data science at Echobox apart from other companies?
Anna Hlédiková: Most tech companies break up data science work into various roles that feed into an overall pipeline. This means that most data science teams are made up of specialists who undertake very specific tasks. At Echobox, data scientists are not specialists. Rather, we get to work on projects that involve all of the duties associated with data science, from data analysis and machine learning operations to optimization. I’d even say the job is sometimes quite similar to that of a software engineer. What really sets us apart from the Engineering team is that we conduct research into the algorithms and we implement machine learning.
Ilaria Gallo: I agree with Anna. To put it simply, our work is divided into two key parts. The first revolves around data analysis. For example, when Echobox releases a new product feature, we’ll measure the impact of this feature on our customers, and we’ll communicate if it has a positive impact on social or email performance. The second part of the role is centered around machine learning, which is essentially building machine learning algorithms that will solve real life problems. Lately, we’ve been working on content personalization for Echobox Email, our newsletter and email software. This new feature allows businesses to send custom emails to each subscriber in a way that the content of the email is completely customized to each individual.
Echobox: What do you like about data science at Echobox?
Anna Hlédiková: Working at Echobox means you have a lot of space to learn and develop your skill set. Though I have a background in AI and had done coding before, I didn’t have a formal background in computer science when I joined the data science team, but I got the opportunity to learn and develop skills related to that field very quickly.
We also have what we call “2-day projects.” Every month, you can decide to spend two days on anything you want to get better at, like learning a new programming language. Lately, I’ve been looking into graph neural networks, as well as attending conferences.
Finally, Echobox is a nice place to work. I didn’t want to work at a big company because it feels like everything is standardized and it can be hard to step out of your role. Equally, I didn’t want to work at a tiny startup because I wanted to learn from a variety of people. Echobox proved to be the perfect fit.
Ilaria Gallo: What makes Echobox special to me is the ability to have your fingers in many different pies. For someone in the early stages of their career, it is incredibly valuable to experience every step of an AI application. You really get to see how each task feeds into the next one, and gain knowledge on things you didn’t think you’d ever care for or participate in.
I also love the atmosphere at Echobox. I wanted to work someplace young and relaxed, where it would be easy to approach people and collaborate on projects. And it’s what I got.
Echobox: What does a normal day in the data science team look like?
Anna Hlédiková: It truly depends on the project we’re working on. I’d say 50% of our time is spent on data science-specific work. This includes researching, writing up investigations or drawing insights from data. 30% is coding and the final 20% has to do with communication and other bits and pieces.
As a data scientist, I’m finding that the skillset is quite different from everyone else in the rest of the tech team. We deal with extremely technical matters, and we need to make sure that whatever we find of importance is easily comprehensible to the commercial teams within Echobox. That way, our Sales and Customer Success teams can easily relay and explain our features, and have a better understanding of how the Echobox AI works.
Ilaria Gallo: Like Anna said, it really depends. There are periods that are research-intensive, and times where we mainly focus on looking for a practical solution to a problem or carrying out data analysis. As we are part of a wider squad, we’ll frequently work with other departments and spend time collaborating. For example, we’ll work with Product Managers to define the specs of a new feature, or have daily chats with software engineers. No day or week is the same in data science!
Echobox: What’s the trickiest part of your role?
Anna Hlédiková: I’d say the trickiest part of my role is having to balance out “being practical” with the need for rigor and thoroughness. This is pretty tough when you’re a perfectionist because you always want to strive to accomplish something very sophisticated, but sometimes our work just needs to suit our clients’ needs and drive quick results.
Ilaria Gallo: We have such a wide range of tasks and responsibilities that it’s hard to say because it really depends on personal interests. Personally, after two years here, I find the research component of the role to be the most challenging, and I’m sure many data scientists would agree. You have to start from scratch to try to find a solution that works for a particular problem, and it’s never a given that your research will actually lead to anything productive immediately. That can be quite daunting!
Echobox: What’s the most exciting project you’ve worked on so far?
Anna Hlédiková: My first responsibility here focused on building content personalization for Echobox Email. The premise of content personalization is to use AI to select the optimal content for each individual subscriber based on what they like to read, and how they interact with certain elements of a newsletter. This means that if a campaign has 10,000 subscribers, there would be 10,000 different combinations of articles and elements. The set up of the problem was very specific, and we couldn’t use out of the box solutions. We had to build a lot from scratch. As my first “real world” project after university, it is very fulfilling to carry out a practical project from start to finish, and witness the entire machine learning lifecycle.
Ilaria Gallo: My favorite thing I’ve worked on so far is the AI-generated share messages feature that is available on our Social product. This feature allows users to save time coming up with messages for their social media posts. In short, it can parse article text to find summary lines or use our language model to generate messages from scratch. There is a state-of-the-art language processing model behind that feature, and we started working on it when that model was in its very early stages. In that sense, our approach to AI-generated content was completely novel, which made working on this project very gratifying.
Echobox: Tell us a bit about your academic background. More specifically, what led you to Echobox?
Anna Hlédiková: I really wasn’t sure what I wanted to study at university. It was either going to be Mathematics or Fashion Design, and I ended up choosing Mathematics based on a gut feeling. I moved to London to study Maths and Statistics at King’s College, where I discovered coding quite late in my second year. I never had considered coding could be part of my career but I ended up loving it so much that I didn’t want to give it up. This led me to apply to a Master’s degree in Artificial Intelligence at Imperial College London, which was the perfect fit as it included coding and mathematics-based modules.
I loved my time at Imperial College, partly because it delivered both a theoretical approach and hands-on method to learning. My Master’s dissertation focused on detecting Alzheimer’s disease from spontaneous speech. This disease can manifest itself through speech quite early on, so I decided to build a machine learning model that could detect Alzheimer’s through text or audio recordings. The main issue was data collection, since acquiring medical data is both expensive and heavily regulated. Building accurate machine learning models requires a lot of data, so a big part of my project involved data augmentation, which is a technique you can use when there isn’t a simple way to collect a vast amount of data. I’ve had the opportunity to attend various conferences this year, even flying to Seoul in South Korea to present my research a few months ago.
Ilaria Gallo: I’ve always preferred scientific subjects, and I knew I wanted to study something related to mathematics at university. However, I also knew I wanted a practical subject, so I landed on Information Engineering. It’s only later in my studies, in my third year, that I discovered data-related subjects. What felt like a natural progression was to pursue this further by undertaking a Masters in Computer Engineering. I focused on machine learning and computer vision, a field of computer science that deals with digital images.
My Master’s allowed me to complete a 6 months internship with Sony, in Germany. This was my first real experience abroad, and it was truly fantastic to work with other interns who came from all around Europe and passionate people who are experts in their field. This is what really led me to look for jobs outside of Italy. Specifically, I was looking for a young company that didn’t have a stuffy, corporate atmosphere – and that’s how I found Echobox!
Echobox: And to wrap up… What’s your favorite dish from your home country?
Anna Hlédiková: Czech cuisine is very heavy on meat. As a vegetarian, that really limits my options but my favorite would be the meat-free version of Svíčková, a creamy sauce made of root vegetables with some traditional dumplings and cranberry jam on the side.
Ilaria Gallo: It sounds cliché, but it’s definitely pizza!
Check out some of the research from our data scientists that’s available to read as white papers, and learn more about what it’s like to work at Echobox. For more behind the scenes at Echobox, have a look at our Instagram.
