When you first meet Sergei, you may find him very calm and pleasant, which he is most of the time. But when it comes to Machine Learning and Foosball, you will be faced with a very different person. We call him Sergio.
Tell us a little bit about yourself.
I graduated with a Master’s degree in Software Engineering and started my career as a web developer. I was good at what I did, and the job was stable, but I wanted more. I wanted more challenges, so I founded a web consulting company and started to expand my experience and expertise into analytics. When the business was doing well, I was not really doing well again. Good amount of business was in a few areas, and I soon started to lose interest because it was not challenging enough. I wanted more complicated and interesting stuff to work on.
What did you then?
I really enjoy transforming a difficult and complex challenge into a generalized, simple problem. I did my research and asked around for advice, and found Machine Learning. Machine learning was really hard in the beginning. I needed to know math well, write codes in different languages, build models, tune hyperparameters, and understand different businesses in order to really apply machine learning and deliver results. This very complicated requirement and process attracted me that I switched my career to a Machine Learning Engineer.
Tell me more about your role now at Ople.
My core responsibility is to understand different machine learning algorithms and learn how to fine-tune them so that they can be used in our system. However, since we are a startup, I do more than a typical machine learning engineer does. At my previous company in Russia, we had a team of machine learning engineers. I only had to do a very small part of a large project. Here at Ople, I have to know and do everything, which is what I love and enjoy.
To tell you a little bit more, let’s think about a typical pipeline for a Data Scientist. Once a project or a business challenge is defined, a Data Scientist would communicate with various stakeholders, come up with a hypothesis. Then, she would collect the data, preprocess it, maybe create new features along the way, and build models. Depending on the target accuracy, she might spend days, weeks, or even months to tune hyperparameters and build different ensembles. This whole process is not a simple task. It requires a tremendous amount of knowledge, experience, and time to successfully finish a project. What Ople is doing is to generalize the process, especially the model building part at least for now. My job as a Machine Learning Engineer at Ople is to make sure that this process works.
What do you do when you are stuck?
My go-to-source is friends. I have a group of friends who are machine learning engineers in both Russia and Silicon Valley, so I ask them when I face challenges. I also spend a fair amount of time on reading research papers to learn new techniques and looking at Kaggle discussion. Since we had Giba, a Kaggle Grandmaster, as the Lead Data Scientist, I email or slack him on product-related specific problems.
What do you like to do for fun?
I like traveling and taking landscape pictures. I enjoy visiting new places, meeting new people, learning new cultures, and trying new cuisines. I just moved to the U.S. last year, so I did not have much time to travel the country yet. I have only been to three states – New York, Nevada, and California – so this year, I want to visit more states and explore.
Do you have a favorite quote?
I like what Steve Jobs said at one of the graduations. He said,
“You can’t connect the dots looking forward; you can only connect them looking backwards. So you have to trust that the dots will somehow connect in your future. You have to trust in something – your gut, destiny, life, karma, whatever. Because believing that the dots will connect down the road will give you the confidence to follow your heart even when it leads you off the well worn path; and that will make all the difference.” (source)
I like this one because it speaks to a lot of what we are doing at Ople. We are developing something new that has no references. We would not know whether what we did was good or bad until we actually do it. I strongly believe that what we are doing is a very innovative and disruptive technology. Because of my confidence, and the team’s confidence, I believe we will be connecting the dots in the future together with big smiles on our faces.