, An Alternative Approach to Becoming a Data Scientist

Data scientist is one of the most prestigious jobs in the industry. With a handsome salary, perks, and reputation, a data scientist is a top career choice among millennials today. Though online data science certifications and courses are good in the way that you can develop in-depth knowledge and/or acquire deeper skills, there’s a much better approach to learning data science that many established self-taught data scientists have followed.

Let’s introduce you to the approach.

1. Solve a real problem

The best trait of a good data scientist is problem-solving. Most employers will assess you on the basis of how well you solve a problem. For this, it is essential to solving a real business problem.

Pick a problem and then solve it using machine learning. There’s no straightforward way to do it. You will need to get your hands dirty.  This will give you real experience and a story, irrespective of your success or failure.

Here are some examples of the problems you can pick.

1. Detect fake news
2. Predict the value of houses in your neighborhood
3. Pet recommendation for people based on their lifestyle

If you succeed at your project, put it on Hacker News, Product Hunt, and/or Github. Add this experience in your resume. If it solves a problem using machine learning, employers wouldn’t care that it was a one-man show.

2. Get a mentor

Several software engineers have successfully transitioned to data science. These folks can guide you better than anyone else.

Reach out to data science professionals on LinkedIn or other professional networking or community platforms to build relationships and seek help. Discuss the problem you are solving and how you can easily solve them. While you are at it make sure that you implement the solutions offered by them and review each potential solution.

Here’s a simple approach you can follow.
1. Text data scientists in your city on LinkedIn
2. Invite them for lunch or coffee.
3. Keep a problem in mind and suggest your solution
4. Once you have implemented the solution, follow up with mentors

3. Do a machine learning internship
Take a part-time job or a short –duration paid assignments where you will get to implement ML. The idea is to get real industry experience, irrespective of how much you make. Focusing on money instead of work will lead you nowhere.

Real industry experience is highly valued and boosts your chances to get your first job fast. If you’re still in school, it would be easy to get an internship. However, if you have a full-time job, doing an internship along with your full –time job would be difficult. So try to squeeze in small-duration AI projects.

You can connect with data science professionals to know about internship opportunities.

4.  Put data science at work in your current job

A more viable option is to think of how AI can help your organization.  You may not have the opportunity to work on AI projects between your 9 to 5 job. If you’re motivated, work on weekends or at night to prove how AI can help solve the program.

If decision-makers find value in your solutions, you may get an opportunity to work on it as part of your projects. This will further go down on your resume where you can elaborate on the details of the project and designate yourself as a data scientist.

5.  Do a data science boot camp

Data science boot camps are a gateway to data science jobs.

Boot camps are costly. They are selective about participating candidates. Often boot camps require advanced degrees like Ph.D. However, not all boot camps are equal.  

Several data scientists have entered the industry via boot camps. Many reputable companies consider boot camps a great way to hire great data, science graduates.

Further, a boot camp will help you in the following ways:

– You will get the opportunity to do consulting for real companies
– You will get to interview with highly reputable data science companies.
– You will get guidance on job preparation.

6. Take a vendor-neutral data science certification

Data science is a broad domain that requires an understanding of multiple subjects including programming, mathematics, statistics, and machine learning. Therefore, it is absolutely necessary to build a strong foundation in these subjects to develop a holistic problem solving aptitude using statistical techniques without depending on one single proprietary platform. Professional vendor-neutral data science certification demonstrates the ability to do so. This is a highly –valuable trait in the industry.

The following are a few other benefits that vendor-neutral data science certifications have.

1. Accelerates job growth for working data scientists.
2. Advantage over non-certified candidates
3. Fast-tracks your entry into your first job

Summing up

Data science career isn’t straightforward like most jobs in the tech industry. Thus, the approach to get a data science job may vary with experience and role. There’s no one-size-fits-all approach. If you’re struggling to get into data science, the above approach is worth the try.