How Can You Distinguish Yourself from Hundreds of Other Data Science Candidates?

A few easy (and not-so-easy) ways to prove to employers that your skills and attitudes place you in a higher bracket.

By Tirthajyoti Sarkar, Adapdix Corp on July 13, 2021 in Career Advice, Careers, Data Science Skills, Interview Questions

comments

Image source: Pixabay (Free to use)

Why bother distinguishing myself?

Because there is a tremendous amount of competition to get a job as a data scientist.

Getting A Data Science Job is Harder Than Ever - How to turn that to your advantage - KDnuggets
Although many aspiring Data Scientists are finding it is becoming more difficult to land a job than it was in previous…

Because there is a mad rush. Every kind of engineer, scientist, and working professional is calling himself or herself a data scientist.

Why So Many 'Fake' Data Scientists?
Have you noticed how many people are suddenly calling themselves data scientists? Your neighbor, that gal you met at a…

Because you are not sure if you can cut your teeth in this or not. Remember that imposter syndrome is well and alive in data science.

How to manage impostor syndrome in data science
What if they find out you’re clueless?

I can go on, but you get the idea…

So, how do you distinguish yourself from the masses? I don’t know whether you can, but I can tell you a few pointers to test yourself. That’s what this article is about.

Ask yourself a few simple questions

Image source: Pixabay (Free to use)

Ask yourself a few questions and count the number of YES answers. The more you have done these the more you are separated from the masses.

If you are a beginner

Have you published your own Python/R (whatever you code in) package?
If yes, have you written extensive documentation for it to be used easily by everyone else?
Have you taken your analysis from Jupyter notebook to a fully-published web app? Or, have you investigated tools that help you do it easily?
Have you written at least a few high-quality, detailed articles describing your hobby project?
Do you try to practice the Feynman method of learning i.e. “teach a concept you want to learn about to a student in the sixth grade”?

At a slightly more advanced phase

If you are not a beginner but consider yourself to be at a somewhat mature stage as a data scientist, do you do these?

Do you consciously try to integrate good software engineering practices (e.g. object-oriented programming, modularization, unit testing) in your data science code at every chance you got?
Do you make it a point to not stop at the scope of immediate data analysis that you had to do but imagine what would have happened for 100X data volume or 10X cost of making the wrong prediction? In other words, do you think consciously about data or problem scaling and its impact?
Do you make it a point to not stop at the traditional ML metrics, but also think about the cost of data acquisition and ML business value?

Building tools and creating documentation: two important skills to have

Image source: Pixabay (left) and Pixabay (right)

Do not spend all of your time and energy analyzing larger datasets or experimenting with the latest deep learning model.

Set aside at least 25% of your time learning to do a couple of things that are valued everywhere, in every organization, in all situations,

building small but focused utility tools for your daily data analysis. Your creative juice will flow freely in this exercise. You are creating something which may not have thousands of immediate users, but it will be novel and it will be your own creation.
reading and creating high-quality documentation related to new tools or frameworks or the utility tool you just built (see above). This will force you to learn how to communicate the utility and mechanics of your creation in a manner intelligible to a wide audience.

As you can see, these habits are fairly easy to develop and practice i.e. they do not need backbreaking work, years-long background in statistics, or advanced expertise in deep machine learning knowledge.

But, surprisingly, not everybody embraces them. And, that’s your chance to distinguish yourself.

How to take advantage of those habits in a job interview?

Image source: Pixabay (Free to use)

Imagine yourself at a job interview. If you did have many YES answers to the questions above, you could have mentioned to your interviewer,

“Hey, check out the cool Python package I built for generating synthetic time-series data at will”.
“I also wrote a detailed documentation which is hosted at MyApp.readthedocs.io website. It’s built with Sphinx and Jekyll”.
“I write data science articles regularly for the largest online platform Towards Data Science. Based on those, I even got a book publishing offer from a well-known publisher like Packt or Springer”.
“Everybody can fit an ML model in a Jupyter notebook. But, I can hack out a basic web app demo of that Scikit-learn function where you can send data through a REST API and get back the prediction”.
“I can help in the cost-benefit analysis of a new Machine learning program and tell you if the benefit outweighs the data collection effort and how to do it optimally”.

Imagine how different you will sound to the interview board from all the other candidates who do well on regular questions of statistics and gradient descent, but do not offer demonstrable proof of all-around capabilities.

They show that you are inquisitive about data science problems.

They show that you read, you analyze, you communicate. You create and document for others to create.

They show your thinking goes beyond notebooks and classification accuracy to the realm of business value addition and customer empathy. Which company wouldn’t love that kind of candidate?

… these habits are fairly easy to develop and practice i.e. they do not need backbreaking work, years-long background in statistics, or advanced expertise in deep machine learning knowledge. But, surprisingly, not everybody embraces them. And, that’s your chance to distinguish yourself.

Where can I get help?

There are so many great tools and resources to help you practice. It is impossible to even list a good fraction of them in the space of one little article. I am just showing some representative examples. The key idea is to explore along these lines and discover helping aids for yourself.

Build installable software packages using only Jupyter notebooks

nbdev: use Jupyter Notebooks for everything

How to make an awesome Python package — step by step

How to make an awesome Python package in 2021

Learn how to integrate unit testing principles in your own ML models and modules development

PyTest for Machine Learning — a simple example-based tutorial

Learn how to integrate object-oriented programming principles in a data science task

Object-oriented programming for data scientists: Build your ML estimator

Build interactive web apps using simple Python scripts — no HTML/CSS knowledge required

PyWebIO: Write Interactive Web App in Script Way Using Python

Write whole programming and technology books right from your Jupyter notebook. Use this for documentation building, too.

Books with Jupyter

Understand the multi-faceted complexity of a real-life analytics problem and how it is much more than just modeling and prediction

Why a Business Analytics Problem Demands all of your data science skills

Imagine how different you will sound to the interview board from all the other candidates who do well on regular questions of statistics and gradient descent, but do not offer demonstrable proof of all-around capabilities.

A couple of things about MOOCs/ online courses

Image source: Author’s own creation

Don’t jump the steps while learning. Follow the steps.

Image source: Author’s own creation

Read board topics and books at every chance

Don’t just focus on reading the latest deep learning trick or blog post about the latest Python library. At every chance, read board topics off the industry’s top forums and good books. Some of the books and forums that I enjoy are as follows,

Image source: Author’s own creation

Summary

Data science and associated skills of machine learning and artificial intelligence are in extremely high demand right now in the job market as more and more businesses adopt and embrace these transformative technologies. There is a lot of competition and miscommunication between the demand and supply sides of talent.

A burning question is: how to distinguish oneself from a hundred co-applicants?

We listed some key questions that you can ask yourself and estimate your uniqueness in some of the skills and habits that make you stand apart from the others. We showed some imaginary conversation snippets that you can have with an interview board showcasing these skills and habits. We also gave a shortlist of resources to help you get started on these.

We listed a couple of approaches for taking MOOCs and suggested reading resources.

Wishing you the best in your data science journey…

You can check the author’s GitHub repositories for code, ideas, and resources in machine learning and data science. If you are, like me, passionate about AI/machine learning/data science, please feel free to add me on LinkedIn or follow me on Twitter.

Original. Reposted with permission.

Related: