“A good first impression can do wonders” – JK Rowling
Gone are the days when people just looked at your CV and decided if you were a suitable candidate for an internship or a job. In the technology sector, people now see the general profile and projects (Briefcase) to shortlist candidates, especially for Data Science. But with time, education companies have started offering paid online projects, so almost all students are now doing projects and internships (paid / unpaid / volunteers).
Therefore, Candidates must not only do projects proactively, but also show your skills to stand out in an opportunity. By showing I mean you must make a mark of yourself. When someone sees your data science portfolio, you must have an exact idea of your interests, previous jobs, achievements and be interested in chatting with you.
Tips for Building an Amazing Data Science Portfolio
1. Have an active Github profile
GitHub allows you to host a remote version of your project where others can see it and even collaborate to form a better version. Always have an active GitHub profile and put the link in your CV. By active profile, I mean you have to work on it regularly because your daily contributions are logged on it and can be seen by viewers. What's more, make sure to do a readme.md (read more about this in Click this) for your profile to personalize your home page.
Here is a sample of my profile to get an idea:
2. Get started with Kaggle
Having a Kaggle account is very important. Not just to show off your skills, but also to practice them regularly. Many companies like ZS Analytics, KPMG, Bain & Co., JPM, etc. have a data science competency like the ones available at Kaggle.
Apart of this, the learning contests available on Kaggle help to understand more about the techniques and tips to apply when dealing with different types of data. Kaggle is also a great platform to show off your skills. You can learn medals and titles (Kaggle 1X / 2X / 3X / 4X Expert, Kaggle Grandmaster) that have a big impact if you put them in the headline of your LinkedIn profile. You can also add a link to Kaggle on your CV.
Below is a random public profile of a Kaggle 3X expert from India,
3. Participate in contests and hackathons
Contests and hackathons help us develop our skills and understand our position in our peer group. Success in contests and hackathons can be listed as achievements that will add credibility to your work.. For instance, the DataPeaker platform can be used to participate in Hackathons. You can also view the top focuses of any competency to learn new and improved approaches.
At the moment, you participate in around 67 hackathons for learning purposes and see the main focuses for previous competitions. They also have Job-A-thons and hire hackathons several times a year, so stay tuned to participate in them.
4. Practice questions using HackerRank
HackerRank is a great platform to improve your Python skills. You have questions that can help you improve your programming skills. along with this, it also offers stars based on the points you achieve for solving those questions correctly. Put (HackerRank 5 stars) in your LinkedIn headline to show your proficiency in Python or Data Structures / Algorithms.
5. Read blogs
Reading blogs keeps you updated on recent developments in the field. They can also be useful in discussions during the interview.. What's more, blogging can be used as a tool to learn new skills. Reading blogs about personal experiences would help you learn more about the industry and what you need to do to find a suitable position in the future.. You can follow Christopher Zita, Analytics India Magazine, DataPeaker Blog / Medium Channel, Towards Data Science (On Medium), KDnuggets, etc. for this purpose.
6. Make your portfolio website
Make a very simple portfolio website. You can code in HTML or use Wix / Weebly to make one. Once you have hosted your website, be sure to include it on your CV too. A website will have a huge impact on a recruiter who sees your profile. It will improve their skills and also give them the opportunity to view their projects and work in the field.. The image below shows a snapshot of my portfolio website.
7. Have a LinkedIn profile
A Linkedin profile is extremely important to everyone. This helps you connect with people around the world who might be working in your field of interest.. LinkedIn also helps you share your work with the community. Many recruiters now use LinkedIn's recommendation system to contact candidates for any vacancies at their companies.. What's more, follow the hashtags in the DS field and # 66daysofdata, #MavenAnalytics (for data visualization), similar a ML.
8. Dor small projects
Get started with projects on well-known data sets like Boston Pricing, Iris, XOR, MNIST, etc. After that, continue to carry out large projects as a recommendation engine, full analysis of some data, etc. Data sets can be found in Kaggle. HR analysis, image analysis, customer segmentation, Netflix data analysis, Uber data analysis are some examples to start projects. Feel free to create your own dataset and then do an analysis.
9. Ddeployment code
Once you have completed a project, try to deploy it on Heroku or AWS or any other cloud platform. This helps you create a comprehensive data science application that can be used. For instance, if you create a movie recommendation engine, soon, using Heroku or AWS, create a website that people can come to, choose the movies they like and its algorithm predicts which movies they can watch based on their interests. This code implementation is very impressive to HR. and can surely help you get an interview.
10. Focus in community building
The above methods are sure to help you build an extremely good profile, but other than that, knowing the opportunities is also important. For it, get involved in the community and build great connections. LinkedIn, Discord, Slack, Telegram are some of the platforms where you can join groups of data scientists who regularly post messages about opportunities you can take advantage of..
I hope you liked this article.
If you have any doubts, leave a comment.
Follow me here, my gray bot will be there to help you.
Media shown in this article on creating a data science portfolio is not the property of DataPeaker and is used at the author's discretion.