Crafting a Standout Data Science Portfolio
What is a portfolio?
Similar to architects and designers, aspiring data-scientist/data-analyst/business-analyst/software-engineer/data-engineer/ml-engineer are judged based on a portfolio.
Think of a portfolio as a bookshelf (Github Profile or website), that contains books (projects) that are filled with words (code).
Understanding how our resume gets reviewed
Here are what people often say about portfolios
Recruiters are the ones who are reading your resumes, no point for portfolios because they are not technical.
This is definitely not true!
When you apply to a job, there are 3 different scenarios you’ll land into.
When the hiring team or hiring manage reads your resume, they will explore every bit of it. They will click on links, check out your LinkedIn, Github, personal websites, published papers, etc. The yellow bit is the moment your portfolio gets reviewed. Notice how all 3 scenarios always get funnelled into that.
What are hirers looking for?
As seniors and working professionals, regardless of data-scientist/data-analyst/business-analyst/software-engineer/data-engineer/ml-engineer, they don’t want to make a bad hire.
They are mostly looking out for 3 main traits:
📖 Adequacy in Technical Proficiency
- Is the candidate technically proficient enough?
- Does he/she already have experience writing production code?
- Can the candidate demonstrate a strong understanding of relevant tools and technologies specific to their role? (e.g., programming languages, software, platforms)
- How well can the candidate debug and solve technical problems?
- Is the candidate familiar with industry best practices and standards?
- Can the candidate adapt to the technical stack and workflows of the organization?
🎓 Ability to Learn
- Does the candidate show a strong willingness to learn and adapt to new technologies and methodologies?
- Can the candidate quickly assimilate new information and apply it effectively in their work?
- How does the candidate approach problem-solving and overcoming technical challenges?
- Does the candidate demonstrate curiosity and a proactive attitude towards continuous professional development?
- How well does the candidate handle feedback and incorporate it into their learning process?
🗣️ Competency in Communicating and Collaborating
- Can the candidate effectively communicate technical information to non-technical stakeholders?
- How well does the candidate work within a team environment? Do they contribute positively to team dynamics?
- Does the candidate demonstrate the ability to listen and understand the perspectives of others?
- Can the candidate manage conflicts and negotiate solutions effectively?
- How effective is the candidate in presenting ideas and persuading others with logical arguments and evidence?
- Does the candidate show empathy and the ability to build relationships with colleagues and clients?
Don't just take my word for it. Here’s what others say about portfolios.
So I recently spoke with some entry level data peeps and I was surprised that only 1 in 10 had a portfolio. Tomorrow I will posting a video on how you can create a data science portfolio for free. I'm hoping you'd get a thing or two from it👍. #DataAnalytics #dataportfolio
— David Effiong (@david_uforo) April 22, 2022A portfolio is not required for a Data Science/Engineering Interview, but it can help a lot
You'll be missing a great opportunity if you do not show one
Some thoughts below
I need 2 Data Science Interns on my team.
Must be good with Python for Machine Learning and have portfolio to demonstrate this.
You should be open to learning and start work immediately.
Location is Lagos and it is Onsite.
Drop your portfolio link in the comment section.
Some examples of portfolio
-
🙋🏻♂️ My portfolio
-
🤏 Minimalist portfolio
-
👨💻 Github-based portfolio
-
📈 Data visualisation portfolio
-
🗞️ Portfolios that uses opensource template
-
🤩 AI Celebrities portfolios
-
🌐 Super good frontend portfolios
Need projects ideas?
There are mainly 5 different types of data science projects:
| Type | Creativity | Complexity | Elaboration | Deployment |
|---|---|---|---|---|
| Creative Hackathon Quality | ✅ | ✅ | ✅ | ✅ |
| Best-Performing-Model Quality | ✅ | ✅ | ✅ | ❌ (Optional) |
| School-work Quality | ❌ | ❌ | ✅ | ✅ (Depends on the project) |
| Pet-project and/or Educational Quality | ✅ | ✅ | ✅ | ✅ (Optional) |
| Professional Quality | ✅ | ✅ | ❌ | ✅ |
| Academia Quality | ✅ | ✅ | ✅ | ❌ (Optional) |
Let’s all aim to have a good variations of data science projects in our portfolio.
Conclusion
Crafting a standout data science portfolio is a crucial step in distinguishing yourself in the highly competitive field of data science. By showcasing a diverse range of projects that demonstrate your technical skills, creativity, and problem-solving abilities, you can effectively capture the attention of potential employers or collaborators.
Remember, a great portfolio is not just about the projects you include, but also how you present them.
Clear explanations, visualizations, and a narrative that ties your work to real-world applications can make your portfolio compelling. Additionally, keeping your portfolio updated, accessible, and tailored to your audience will ensure that it continues to serve as a dynamic representation of your evolving skills and interests.
As the field of data science continues to grow and change, your portfolio will be an essential tool in navigating your career path, opening doors to new opportunities, and establishing your identity as a data scientist. Let your portfolio be a reflection of your dedication to the craft, your passion for discovery, and your commitment to making an impact through data science.