15 February 2023
If you don’t want to read a long intro - feel free to click on table of contents to jump further.
And Part II is here
First of all, this post is not GPT generated. The pictures are a credit to midjourney, but the writing is not. It could have been, though, but I decided to do it myself to develop my writing habit. I’ve enjoyed writing, and at some point, I thought a journalist career was ahead of me. Still, my teachers and my peers gently guided me on another path - math, physics, alcohol consumption, actuarial science, and eventually, data science. I’ve been trying to keep writing about topics of different importance but all of these scribblings went into a desk drawer.
On February 24th, russia invaded my country. I was on a business trip, so that I couldn’t return home safely. I recognize my fortune in being able to stay abroad as there are thousands of people who would envy my situation. But at the same time, during the first year since the invasion, I’ve lived in 6 different apartments in 5 countries.
With all this stress, I forgot about my usual duties - family, and work. I quickly noticed that my productivity had diminished dramatically. And most of my habits were gone! I couldn’t concentrate, was reading the news and discussing it with friends and colleagues, and I procrastinated the rest of the time. After five months, I decided to take back control of my life. I’ve asked myself - if I am starting from this low, where should my priorities go? Where should I put my efforts first, and in which habits and processes? I analyzed the essential habits of the data scientist and decided to share my findings. I’ve split them into technical and non-technical categories for more convenience. But first things first.
There is a necessary prerequisite that people always forget about and keep wondering why after perfectly following somebody’s advice, they still come up short of their goals. I am an adamant advocate of the view that knowing yourself and understanding how your mind and body works are of utmost importance before trying to prioritize and develop your habits.
Everybody is different. Some wake up early to tackle most tedious tasks even before breakfast, and some are insanely focused around midnight and tend to work until 2-3 am. Some embrace chaos and are most creative when a routine does not restrict them, and others need to have a day planned by the hour to achieve peak performance. If you look around for successful people that you know, you will see that their success is not tied to some specific habit but rather to the ability to dedicate a lot of time to the task at hand. That is why I want to emphasize that those differences are neither good nor bad in isolation, but you need to know how your brain and body function in different circumstances. How stress resilient are you? Which time of the day/week/year can you perform better or worse? What level of supervision do you need? What environment helps you concentrate better? So you need to know what drives you, brings you to a “flow” state, and motivates and entices your curiosity.
Now, the fair question is what to do next if you don’t know all that. Well, you should figure that out by putting yourself in different situations and being mindful of what is going on with you. You are a data scientist, right? It is all about experimentation and some of your hyperparameter tuning. Knowing yourself is crucial to a data scientist, as any other person who wants to achieve excellent results in what they are doing. But besides that - here are some habits more specific to data professionals.
Maybe you are already getting paid for reading research papers. If yes - good for you, you can have this checkbox crossed out. For others, I want to emphasize the importance of doing so. Yes, not only papers but blogs as well, maybe even watching some videos will do the trick for you. The main goal is to stay connected and interested.
Recent advances in content generation are really inspiring, and it is good to understand, in general, how they work. But if you are not in that business just yet, there can be tons of materials for various topics like interpretability, model bias and, fairness, MLOps. There are tons of talks, courses, workshops, tutorials, and other ways to engage. If you are like me with tons of browser windows opened, then start closing them one by one each day.
One of the questions I often get is how to start. Where can one consistently find the materials? This is where recommendation algorithms of social media networks may finally become helpful. You can select one of your social networks (like Twitter), unsubscribe from everybody, and subscribe only to some data scientists or leaders in the field. Then the platform will show you the content they like and recommend some other folks in the area, you can also follow them. You aim to create a perfect bubble, and, again - this is one of the few things these platforms are really good at. Don’t engage with any non-data science-related content on this platform. Different platforms have a different feedback loop, but if you see something not related to data science - feel free to flag this content as the one you’d like to see less. Besides Twitter, Youtube can also be an excellent place, and many proponents use LinkedIn for this purpose.
How it helped me?
External links:
https://fullstackdeeplearning.com/ - MLOps course for DL, all teachers are active on Twitter
https://twitter.com/goodside - one of first Prompt Engineers. If you wonder how to wtite good prompts for Generative AI - he was one of the first to start doing that professionally.
https://www.youtube.com/c/nicholasrenotte - does some live coding sessions and speedruns interesting use cases
It is good practice to notice what you have done this week or month. If you or your manager set your OKRs (new fancy word for goals), then when the time comes to check their achievement, it may be hard to remember all the details of the project you did, for example, two months ago. So you will have to spend a lot of time finding specific information or metrics you wanted or agreed to achieve. In contrast, keeping notes like Wins of the week or Wins of the month will help you immensely.
Not only that - but depending on how the compensation review in your company is performed, you can talk to your manager and justify your raise. Besides that, do you remember the mantra of all the articles about constructing your CV? Like you have to mention the goals you have achieved. If you are not appreciated enough in your current job, then some of these notes can be used in your CV to attract potential employers. If you are looking into a particular framework how to exactly keep track of your wins then here is a good article on how to present your wins yousing STAR (Situation, Task, Action, Result) method
How it helped me?
External links: Apparently there are not a lot of materials on this habit somehow. Maybe others don’t feel it is important to do so, or maybe it doesn’t matter how you do it, if you do. But here is one good overview of the topic by the link below
https://www.indeed.com/career-advice/career-development/how-to-track-accomplishments
Your calendar can be your ally and can be your enemy. I have seen people thinking after joining a new company thinking, they should fully adjust to the calendars of their colleagues and make themselves available for calls and meetings 24/7. Instead, balancing corporate culture and your free time would be better. Don’t worry about missing out on some important meeting, let the owner of the meeting figure out how to accommodate everybody - worst case scenario, they will ask you directly if the timeslot at 2 am your time works for you.
Well, technically, it is not a habit - mastering your calendar is a skill, but it helps you develop some good habits. Data scientists often require uninterrupted work time, and they don’t get it because somebody from the product tries to reach you out for a “quick call” or a colleague asks for help “for 5 minutes”. Before Covid, the calendar was not that relevant, but now, with many working from home, it has become the first place where most people will look if they need your time. Be careful with your time, cherish it, hold it close, and you will be rewarded.
Here are some tips on how to master your calendar:
And remember - some of the meetings should’ve been just emails.
How it helped me?
External links:
https://www.calendar.com/blog/scheduling-skills/
Projects sometimes take months, from the start to actually take off and to the final delivery. Sometimes projects become dormant for a while to be resurrected at the least convenient time. And now you have to remember what was already discussed, why particular decisions were made, and should we even proceed further. Countless times I have seen project discussions resurface after hours of meetings just to figure out that nothing really has changed and there is no point in reviving it.
Taking notes is a habit where doing wrong is way better than doing nothing. For example, I am taking notes by hand on paper in a physical notebook, and it is still hard to find something in it after several months. But it is possible. And then you have your whole history of communication at hand, so you will not waste time trying to remember some details from the past.
So long story short - ensure that somebody is taking notes at important meetings - either you or delegate it to somebody else if you are, for example, presenting or leading the call.
How it helped me?
External links: https://jerrynsh.com/the-best-thing-ive-done-for-my-career-is-taking-notes/
There are also many videos that promote some tools like Notion, Obsidian etc, you can search for them on Youtube.
Next is onto technical habits - Part II
