- GRATITUDE DRIVEN
- Posts
- Escaping the Experience Trap
Escaping the Experience Trap
How to get your first data science/ML job (with no previous experience)
Marina Wyss
November 21, 2024
Welcome to Gratitude Driven, a weekly newsletter where I share practical ideas and insights across personal growth, professional development, and the world of AI and data science.
In Today’s Newsletter
Escaping the Experience Trap
Breaking into data science or machine learning can be challenging – especially when every “entry-level” role seems to require years of experience.
As a newcomer to the field, you won’t be able to demonstrate your skills through previous job titles, so you’re going to need a portfolio. But the thing is, not all projects are equally as impactful.
Here’s a breakdown of portfolio project types, from least to most powerful:
1. Follow-Along Projects – (Least Effective)
Step-by-step tutorials on platforms like Coursera or Udemy are great for learning the absolute basics, but they won’t impress recruiters much. These projects mostly show you can follow instructions, rather than demonstrating independent problem-solving.
2. Certificate Capstone Projects & Academic Work
While these projects reflect some level of independent thought, they’re often generic and limited in scope. They add value but won’t make your portfolio stand out.
3. Kaggle Competitions
Kaggle is popular, and a high ranking can be impressive. However, Kaggle datasets are typically small, clean, and well-structured, which doesn’t reflect the complexity of real-world data. Recruiters have seen these projects often, so they rarely stand out.
4. Open Source Contributions & Stack Overflow
Contributing to open-source projects on GitHub or actively answering questions on Stack Overflow highlights initiative and a willingness to tackle real-world issues. These activities demonstrate that you’re learning from and contributing to a community of experts and that you can handle troubleshooting and problem-solving independently. Helping resolve issues in libraries you use regularly can be a solid addition to your portfolio.
5. Self-Motivated Projects – (More Effective)
Designing projects from scratch and working on them end-to-end shows employers that you can identify problems, source data, and develop solutions. Choose a problem you’re genuinely curious about, use public data or APIs, and build end-to-end pipelines that refresh with new data over time. These projects highlight initiative and critical problem-solving skills and will teach you much more than simply following someone else’s project design.
6. Real Projects for Real Clients – (Highly Effective)
Working with real clients—even as a volunteer—is the best way to gain relevant experience. Not only will you handle actual data, but you’ll also learn about the industry’s real-world challenges and needs.
If possible, look for opportunities within your current role, even if it’s not a technical position. For example, if you work in retail, consider creating a demand forecasting model for product inventory, or develop a dashboard for customer feedback or sales trends.
Completing a project for your current job might even help you transition to a role closer to your goal. While you may not jump directly from Admin to MLE, a path like Admin > Financial Analyst > Data Analyst > Data Scientist > MLE can become viable.
If you’re unemployed or can’t find opportunities in your current position, reach out to small businesses or non-profits and offer your services for free. Since they may not be familiar with DS/ML, invest time in understanding their business and explain the potential impact of your project.
All these approaches provide real experience, expand your network, and could lead to paid work in the future.
Check out this video for more examples, and tips on catching the attention of recruiters and hiring managers once you have a solid portfolio.
Do You Just Fit In, Or Do You Truly Belong?
Lately, I've been reflecting on the difference between belonging and simply fitting in. I’m lucky to be able to get along with most people and blend into various social circles, but the feeling of truly belonging is rare. I can count on one hand the people I feel completely comfortable being myself around.
That’s something I’m hoping to change with Gratitude Driven. I hope to find and connect a community of people who are driven, purposeful, joyful, and motivated.
I encourage you to take a moment and consider whether you feel a sense of belonging where you are in life. Can you be yourself? If not, why? What part of you is waiting to be expressed, but hasn’t had the chance to come out?
Here are some other questions to think about:
What qualities make you feel at ease with others? Are they present in your current relationships?
When do you feel most authentic? Are there people or spaces where that feeling naturally arises?
Are you compromising any values or desires to fit in? How would it feel to let those compromises go?
Who truly "sees" you—your strengths, your quirks, your dreams—and do they celebrate them?
What environments energize you? Are there communities or groups that embody this energy?
If you could bring more of yourself into your interactions, what parts would you reveal?
What I'm Reading
I recently finished In My Time of Dying by Sebastian Junger. This account of a near-death experience and what that means for us in this life was beautifully-written and thought-provoking.
Forwarded this email? Sign up here.