In the author's spotlight series, the TDS editor chats with community members about their career paths about data science, writing and sources of inspiration. Today we are excited to share our conversation Egor Howell.
EGOR is a data scientist and machine learning engineer specializing in time series prediction and combination optimization. He runs a content and coaching business that helps people infiltrate data science and machine learning and teach technical topics.
Let's start from the beginning: What has sparked your initial interest in data science, especially since then? You did not follow traditional CS degrees or boot camp routes?
I can almost prove my career to Deepmind's Alphago documentary on my own. I was very interested in machine learning and the possibilities of solving virtually any problems. After that, I was looking for a career using machine learning, and of course there were data scientists out there. So, since then, I have basically self-defined to be one!
You wrote We conduct over 80 data science interviews. What were some important insights you gained from that experience, both about the employment process and your own growth?
Interviews are skills and are very different to what you do at work. It's basically a game and you have to learn how to play it like you do almost anything in your life.
The central insight is that you basically have to prepare. I was shocked by how many times the candidates really don't even know what the business is doing!
Another important point that people overlook is soft skills and intangible assets. Unfortunately, let's say someone is very monotonous and shy, but knows a lot. In that case, they are less likely to get a job compared to people who are charismatic, friendly and generally bring good energy.
And finally, don't talk for more than two minutes at a time. I interviewed people who spoke, spoke, spoke. If you notice that you've been talking for a while, say, “You can go into more detail if you want.” In this way, the ball is on their court and they can move forward with the interview if they want. There's nothing worse than someone who keeps talking, as the interviewer doesn't allow them to ask all the questions. What's more, it's a skill that allows you to explain yourself concisely.
One of your more provocative articles is entitled “Stop building useless ML projects. ” Why do you think so many portfolio projects are missing out on the mark? Why does the project really make an impact?
People are constantly looking for shortcuts and don't want to spend time thinking about quality projects. An impactful project is personal to you and will take at least a month to solve problems and answer questions you want to know.
There is no secret. It's about the efforts people don't want to put in most of the time. In that specific post, there is a framework that people should follow if they want to find an impactful project for themselves.
You often write with a clear audience in mind: career switchers, beginners, aspiring ML experts. How do you decide what to write and who you want to help the most?
It was tough at first, but now I'm asking the audience, reading the comments and seeing what people are looking for.
My goal is to help people get into the field, but I am cruelly honest along the way and don't sprinkle anything sugar.
In most of my posts I often say “I promise nothing” and how difficult it is, and that may not be the right job for anyone.
What is it I surprised you When you start working full-time as a machine learning engineer, do you want to know that more people will come in?
In contrast to model development, you spend a lot of time maintaining your model and infrastructure. Work doesn't stimulate 100% of the time.
You've published a lot of career advice – from work preparation How to highlight your DS portfolio. Did writing regularly shape your own thoughts or your career path?
You are thinking about writing, so the better you write, the better you will think. What people don't tell you is that a lot of work writes. You write plans, documents, tickets, etc. This skill is important. Because if you can clearly describe yourself, it will go a long way in life.
What trends do you have in machine learning and AI? How do these trends shape your focus and ambitions?
I really “hate” AI. I think it's overrated, but I'm definitely not going to take over the job for at least the next five years. Personally, I think it's a “bread flash,” so I don't really try to learn that. I would like to focus on areas that have existed for decades, such as statistics, operational research, and time series.
For those who feel stuck, they may be in the role of data analysts and are they struggling to break into ML?
Take everything one step at a time and don't try to think too ahead. First, focus on the project, then the resume, then the application, interview, and then the offer negotiation.
If you're not getting anything, there's no point in focusing on the interview. It's best to spend more of your time on your resume and projects. A single focus is how you can make progress.
To learn more about Egor's work and stay up to date with the latest articles, follow him on TDS, YouTube and LinkedIn.
