7 Regrets From My First Year As A Data Scientist: What I Wish I Knew Before I Started My First Data Science Job

Starting my data science journey in September 2021 was exciting, but reflecting on that first year, I realize there are many things I wish I had known. Here are seven key regrets, along with actionable advice and examples to help you avoid the same pitfalls.

1. Learn Git

Regret: Underestimating the importance of version control. Action: Invest a week to learn Git basics. Use resources like freeCodeCamp’s one-hour video tutorial. Create a GitHub account and start adding projects to get hands-on experience. Version control is essential in team environments and showcases your work portfolio. Example: During a project, my colleague and I ended up overwriting each other's code because we weren’t using Git properly. This could have been avoided if we had used Git branches and commits.

2. Don’t Only Use Notebooks

Regret: Relying solely on Jupyter or Google Colab notebooks. Action: Get comfortable with developer-focused IDEs like PyCharm and VSCode. Implement some projects using these tools and familiarize yourself with unit tests, linters, and package managers. This knowledge is crucial for deploying machine learning models in production. Example: My first deployment failed because the code wasn’t properly modularized or tested. Using PyCharm helped me write cleaner and more production-ready code.

3. You Can’t Learn Everything

Regret: Trying to master all areas of data science. Action: Focus on one domain that aligns with your role or industry. Build deep knowledge in one area for a year and then pivot if needed. Remember, your career is long, and there's ample time to explore different fields. Example: I tried to learn deep learning, natural language processing, and computer vision all at once, which led to burnout. Once I focused on NLP, my expertise and job performance improved significantly.

4. It’s More Than Algorithms

Regret: Believing data science is all about deploying fancy algorithms. Action: Emphasize data exploration and understanding. Most of your work will involve manipulating and analyzing data rather than just building models. Always prioritize data over models. Example: I spent weeks fine-tuning a model but realized later that simple data aggregation could have answered the business question more effectively.

5. Not All About Being Technical

Regret: Thinking technical skills are all that matter. Action: Develop strong communication skills. Be able to articulate your findings to non-technical stakeholders. Focus on how your work impacts the business and drives decisions. Example: I presented a technically complex solution to a non-technical team, and they were lost. Learning to simplify and focus on business impact made my work more appreciated and actionable.

6. Dealing With Ambiguity

Regret: Expecting clean, pre-defined datasets. Action: Get comfortable with ambiguity. Learn to frame business problems in a data science context, gather relevant data, and find solutions. Work on more data gathering and cleaning projects to build this skill. Example: I was once given a vague business problem without clear data. Learning to ask the right questions and gather the necessary data made it possible to provide a useful solution.

7. Stop Rushing

Regret: Rushing through projects without thorough validation. Action: Take your time to ensure accuracy in your data and analyses. Triple-check your work to avoid errors. Deliver high-quality work rather than rushing to meet deadlines. Example: I hurried through an analysis and ended up joining tables incorrectly, leading to a wrong conclusion. Spending extra time on validation would have prevented this mistake.

Summary & Further Thoughts: Transitioning to a data science job from studying is challenging, but you can ease this process by learning from the regrets listed above.

Another Thing! I share weekly tips for becoming a better Data Scientist and the latest AI news in my newsletter.Subscribe for pure actionable insights from a practicing Data Scientist. No fluff or clickbait, just valuable content to keep you ahead.

#DataScience #CareerAdvice #FirstJob #MachineLearning #Git #IDEs #DataCleaning #CommunicationSkills