Youssef Hosni, a senior data scientist, shares insights on how to build a data science portfolio and discusses his experience in starting and running a podcast. He emphasizes the importance of having a portfolio to showcase your skills and compensate for a lack of industrial experience. Youssef provides practical steps for finding project ideas, conducting data analysis, and documenting your work. He also offers advice on starting a podcast, including the challenges and feedback he has received. He emphasizes the importance of starting and taking action, even if the quality may not be perfect initially. Youssef also highlights the value of learning from guests and their diverse experiences. He encourages continuous improvement and iteration, acknowledging that the first version may not meet expectations - this is an expected part of the process.

Takeaways
- Building a data science portfolio is crucial for showcasing your skills and compensating for a lack of industrial experience.
- To find project ideas, identify domains of interest, research case studies, and explore companies active in those areas.
- When starting a project, focus on complex data sets to demefonstrate your data cleaning, feature engineering, and modeling skills.
- Proper documentation, including an introduction, data description, modeling approach, and code documentation, is essential for a portfolio project. This part can really make you stand out in the job market.
- Embrace continuous improvement and iteration.

Where to find Youssef
LinkedIn: https://www.linkedin.com/in/youssef-hosni-b2960b135/
Substack: https://youssefh.substack.com/
Kaggle: https://www.kaggle.com/youssef19
GitHub: https://github.com/youssefHosni
X: https://x.com/Youssef70125494?mx=2

Where to find Ida

Website: https://idajohnsson.com/
LinkedIn: https://www.linkedin.com/in/ida-johnsson/
X: https://twitter.com/IdaBJohnsson

Timestamps

00:00 Preview
01:01 Intro
03:05 The importance of a data science portfolio
05:02 The value of sharing written content
09:35 Creating a Data Science Portfolio 101
16:15 Combining your interests with a specific problem
18:45 Don't get stuck in the research phase
21:43 Stand out in the crowd
29:53 Model deployment
30:41 Good documentation
34:55 Youssef's podcast
37:38 Start doing instead of perfecting an idea

Reach out to me at www.idajohnsson.com