A seasoned Data Scientist updates and maintains his portfolio regularly. 

A Data Science Portfolio serves as a reliable repository for projects and data-sets worked upon by Data Scientists. In this article, we will discuss exactly how helpful a Data Science Portfolio is and how to construct the perfect portfolio for landing desirable jobs in 2020.

 

Learning from the Expert

 

The best way to go about doing anything is learning from the experts. They will tell you the most important part- “How not to do a particular thing?” When you learn the crucial and key aspects from the masters of the field, the success rate is much higher.

 

In this feature, Manvender Singh, the Chief Data Science Mentor at Accredian and top Data Science Academician in India shares the importance of building a great Data Science portfolio and what not to do in the process.

 

Building a great Data Science portfolio is extremely critical to get a Data Science job. What I’ve seen is that a lot of students, who are learning Data Science, somehow underestimate what Data Science portfolio can do for them. 

 

In this article, we are going to discuss, how to build a Data Science portfolio that will get you a Data Science job. 

 

Why Have A Data Science Portfolio?

 

Let’s approach this question from a recruiter’s perspective and turn the tables; let’s see what a recruiter’s challenge is?

 

A recruiter’s challenge begins when a Data Science job posting for a typical role gets, for instance, 500 applications; most of which are irrelevant. Among the relevant candidate applications, a recruiter finds most applications stuffed with all kinds of buzzwords, like machine learning, decision tree, names of several machine learning algorithms, deep learning, etc.

 

The biggest challenge for a recruiter now stands tall-  whether such candidates have some proven ability to work on these algorithms or tools, or the candidate has just done keyword stuffing, i.e., candidates have just put all kinds of keywords in their resume.

 

This is where a portfolio comes into the picture. 

 

A portfolio gives a demonstrated proof of your experience in Data Science to a potential recruiter.

 

The recruiters want to see a demonstrated experience and not just keywords that are stuffed in a resume; this is why a Data Science portfolio is required. 

 

Essentials of an Exceptional Data Science Portfolio

 

Now that it is clear why a Data Science professional should have a portfolio, I will now discuss at length, the components that go into making a successful portfolio. 

 

I will take a do’s and don’ts approach to discuss these components.

 

What to DO?

 

Here are the three do’s, the point-of-actions for you to take up immediately, to build a successful portfolio.

 

1. Choose Your Projects to Include in the Portfolio

 

The first point is selecting your projects.

 

There can be two ways to do it:

 

– An exhaustive collection of all your work, in the form of a GitHub profile
– A collection of your best work 

 

Instead of writing all the small projects in your GitHub profile, mention the best project, most complex project and most interesting project that you have worked on.

 

This will help the recruiter to take a quick decision. Now, let’s look at the pros and cons of both the approaches. 

 

For instance, you are trying to build a Data Science portfolio in which you are detailing your complete work done till now; all the projects completed so far.

 

The recruiters will be confident that you’ve done some kind of Data Science but they will not have the time to go through all the projects. This might leave them in a fix; which project to exactly interview on. 

 

To avoid this, here’s what we recommend at Accredian. 

 

Work on a Data Science portfolio, wherein you are featuring the best of your projects.

 

For instance, projects like Titanic data-set and IRIS data-set, in what you claim to be a Data Science portfolio, will not leave a good impression on a potential recruiter.

 

But if you are taking some complex data-sets or some industry problems and using Data Science to find a solution, your portfolio will surely stand out as a genuine one.

 

2. Select a Platform to Feature Your Portfolio

 

The second point is about the selection of the platform. Different people feature their portfolios in different ways.

 

• Candidates build their website with their own names in the URL and feature their project 
• Candidates build a GitHub profile and upload their project 
• Candidates blog about the projects that they have done. For each project, there is a different blog post and then there is a master blog post, which has a link to all the blog posts.

 

All these three approaches are good and you can choose from any one of them. Here is our order of preference to feature a portfolio:

 

• In the first place, I would recommend that if you have a GitHub profile, don’t build an entire website or platform from scratch.
• Next in my order of preference is building your own personalized website; it does not take a lot to build a blog or a website. 
• The last in my preference order is if you can blog about your projects.

GitHub is the most widely recognized platform for showcasing your technical prowess. It has a myriad of built-in functionality.

 

3. Elucidate Your Projects

 

The third do point for you is to describe your projects well.

 

I see a lot of portfolios, wherein students have just uploaded the data-set with a little bit of what they did, to go with it. In the Data Science world, you need to describe the data-set at length with all the columns and variables, etc. 

 

These were the three do’s that will help you meet your goal of making your portfolio and projects come alive in front of a recruiter.  

 

What NOT TO DO?

 

We will now discuss the three steps you should never take, if you wish to have a portfolio that speaks for you.

 

1. No Trivial Projects in the Portfolio

 

First thing in this sequence is, as I said, don’t include insignificant projects in your portfolio; they will end up making you look like a newbie Data Scientist rather than someone who understands Data Science in-depth. 

 

2. Start Working on Projects with the First Step in the Data Science World

 

The second recommendation I have is that you should constantly work on projects, right from the initial months of starting your Data Science journey; what you should not do is, wait for 6 months; you will upload your project, once you have learned machine learning because now you will start dealing with that process. 

 

Start immediately!

 

The moment you have something substantial or tangible that you’re working on, start uploading it on GitHub.

 

3. Don’t Clone Other Projects

 

Do not copy anyone else’s project. 

 

This is also a common practice we see in a number of portfolios. Students end up copying other’s projects and when they are called for an interview, they don’t know anything about those projects.

 

With the help of a very few questions, a recruiter is able to understand that the project in front of him is a clone and nothing more than that.

 

Finally, the last step for building a successful Data Science portfolio that will get you to interview calls for sure is to have clarity about how to position yourself in an interview for a Data Science project. 

 

To make this step successful, you need to be prepared for interview questions like:

– Which is the most difficult project you have worked on?
– Which is the easiest project you have worked on?
– Have you worked on a problem statement or a project which uses this kind of machine learning technique? 

 

Therefore, while preparing for the interview, you should go through your projects, review them well and have stories around it.

 

This will be the final approach that will bolster the chances for your portfolio to get recognized.

 

These are the steps to build a great Data Science portfolio that will reward you with a successful Data Science career. But remember, even when you have cracked the interview and got the job, keep on revisiting your portfolio and add great projects and insights.

 

Do let me know if you wish to seek any advice on creating portfolios or have any doubts. Feel free to comment in the comment box below.

 

All the best!!