While there is a lot of hype about Data Science being the sexiest job of the 21st century, not much is said about the Data Science journey to acing the job.
What do you need to land a job in the next 6 months?
Is a Data Science Certification sufficient to crack your job interview?
What should you include in your CV?
To land a good job, enhance your Data Science Certification with a good Data Science portfolio. Together, they will help you ace your Data Science job interview. Whether you are a fresh graduate or a professional changing career direction, a Data Science portfolio helps you stand out against other applicants. The portfolio showcases your critical thinking and communication skills, as well as problem-solving abilities. It is relevant to all Data Science interviews regardless of the Data Scientist job role or industry.
What is a Data Science portfolio?
A Data Science portfolio is the public evidence of your Data Science skills. Projects executed during internships or capstone projects are visible only to you, your teammates and your teacher or mentor. However, putting out your project in the open domain together with write-ups and discussions, are a proven and validated testimony to your own skills. A Data Science portfolio includes details of projects you have worked on together with the links and blogs that exhibit your knowledge base. You may even want to publish your research in a journal or a leading Data Science platform or share the same at a major event or conference as a speaker. Contest participation, research, projects, and publication; together make up your Data Science portfolio.
Why build a Data Science portfolio?
Most companies look for applicants with real-world experience, whether for an entry-level job or an advanced job role. However, if you are a Data Science newbie, how do you go about building an experience? It is certainly a catch 22 situation, and the best way to overcome this challenge is to work on your own Data Science projects. Very often, a Data Science enthusiast will receive a job offer out of the blue based on the project at Github or an answer to a particularly sticky problem at a Q&A site like Stack Overflow. Having a public portfolio means a higher probability of someone seeing your work and contacting you with a job opportunity. It is equally probable that the interviewer has come across the work you have done if you are active on some of the leading platforms and blogs.
One often forgets that software engineers and data scientists also Google their issues. If these same people solve their problems by reading your public work, they may also choose to reach out to you.
Projects are perhaps the best substitutes for work experience. So if you don’t have a Data Science related work experience, the best option is to work upon a Data Science project and leverage that as part of your Data Science portfolio.
How do you start a Data Science portfolio from scratch?
Crawl through leading job sites advertising jobs in Data Science roles. Zoom in to a handful of Data Science jobs that interest you, or jobs you would take if offered to you. Make a whitelist of the skill sets, tools and languages mentioned in the job responsibilities.
Prune the whitelist. Retain the list of skills, tools and supporting languages which are common to the jobs that appealed to you.
How many of them do you have or know? Strike them out.
The remaining result on your list is your to-do list. Learn these technical skills and master the tools. Make them part of your portfolio.
Generate about a handful of project ideas, typically 2-3 more than your target number. Look up the blogs and platforms listed here on our Springboard India website. Take your time over project selection, maybe discussing it with your mentor. Identify projects that reflect closely the job responsibilities of the positions you are interested in.
The biggest hurdle is usually the dataset. Try to get free-to-use datasets that closely mirror the Indian business ecosystem or those that allow you to flex your technical and problem-solving skills. The problem of finding an ideal dataset for your project type is the reason why you may have to drop some of the initial project ideas. However, aim for 3 projects (or more, if you are willing to give the time).
However, do not over-think. Begin on the first project. The lessons you take away from your first task will help you re-evaluate and decide on how to go about the second one. As you keep working through, you have a clearer idea. So for your next projects, you may even end up selecting a dataset or projects that had not occurred to you before.
Get cracking on your projects. Do a structured write-up for each of the projects. This is as important as the problem-solving itself because your ability to explain the methodology and reasoning is the ultimate job clincher. Project write-up demonstrates your skills. So take your time over it, and address questions such as these.
- What is the question or problem you are looking to answer?
- Why is it important?
- What data did you use?
- What was your data source?
- How was the data sampled?
- How did you model the data, and why?
- What code did you use?
- How did you fit the model, and validate it?
- How did you conclude that the results make sense?
- Did you visualize the results?
- How to communicate the results to your team or seniors?
- Could this be done differently? How?
- What are the logical steps to take forward this project?
Your portfolio opens the doors to a Data Science career
A. Create your project
We have already covered how to start your project from scratch if you are fresh in Data Science. However, if you are a working professional or Data Science enthusiast, you may want to skip the above and head straight to the project creation stage.
B. Build a project footprint on the web
Put out your project for the world to see. You would be surprised at the wonderful feedback and suggestions you receive. Many a time you look at a problem from a whole new perspective. Or you may discover the rationale behind a suggestion to use another method or tool. It is a great community out there, and putting out your project helps you improve upon both, the project and your skills. Explain your method or sound your ideas. Keep improving your projects as you move along.
You can post your projects on GitHub, take part in Kaggle competitions, or fill out descriptions for your GitHub repositories.
- Display links to your projects or ranks in competitions you have participated in, on your CV and professional profile on LinkedIn.
- Common projects do not distinguish you from other applicants. So include novel or currently relevant projects.
- Keep fine-tuning even through your job hunt.
- Keep on revisiting. You will discover there is always room for improvement.
B. Talk about your projects
Data science is about sharing, problem-solving, troubleshooting, fine-tuning codes, and much more.
Sometimes, it is good to blow your own trumpet, and this is most true in Data Science projects. Blog about your projects. Post links and talks about them. You have invested plenty of time and effort into your projects, and now it is time to take pride. Blogging is one of the most effective and robust forms of self-promotion, as it helps you explain the process adopted and the reasoning behind the use of methodologies.
All Data Scientists have begun from the bottom rung of the ladder someday. It is understood that initial projects will lack the sophistication that projects by experienced data scientists have. However, by continually working on projects and actively engaging in discussions on the same, you keep getting better and better. Your projects focus on tougher problems than before and relate to more relevant and current scenarios.
As you keep practicing on your projects, your Data Science portfolio just gets better.
So if you want to land a job in the next few months and have recruiters reach out to you with employment opportunities based on your portfolio, work on enhancing your portfolio with complex and multifaceted projects that will showcase your abilities.