- Looking to transition into data science? Here are 5 paths for a non-data science person to land a role in this space
- The 5 backgrounds we cover in this career transition article – software engineering, finance, UX, application development, and a non-technical fresher
- I’ve provided links to plenty of resources and learning paths to help you start your data science journey
Are you looking for a role in the data science space? You’ve come to the right place!
It feels like half the world wants to move into data science these days, with spectacular perks and a plethora of openings on offer in the industry. Organizations are investing heavily in data science talent to stay or move ahead of their competitors. As a data science aspirant, you couldn’t have picked a better time to change your career!
But this comes with its own set of challenges. I’m often asked by folks in my network about how they should transition into data science. People from all sorts of backgrounds – IT, Sales, Finance, HR, Healthcare, etc. – they all want a piece of the data science pie.
Let me first put your doubts to rest – it is entirely possible to transition into data science from your current line of work (or study). And that’s what we’ll talk about in this article!
I’ve collated my pick of the top answers from around the internet in this article. We’ll be talking about how to make a transition into data science from these backgrounds:
- Software Engineering/Development
- UX (User Experience)
- Application Development
- A Fresher with no Relevant/Technical Background
I would also strongly recommend checking out our comprehensive, structured and free learning paths for 2020:
- A Comprehensive Learning Path to becoming a Data Scientist in 2020
- The Ultimate Learning Path to Master Deep Learning in 2020
Let’s dive in!
This pick is for the software engineers out there looking for a transition into data science. Yassine has listed down the things you should do to get into data science. He also explains what kind of roles would be suitable for a software engineer getting into this field.
Answer by – Yassine Alouini
To start with, if you really enjoy software engineering, then you should consider becoming a data engineer or a machine learning engineer.
These aren’t technically speaking “canonical” data scientist roles, but close enough and considered among the larger data science fields.
If you would still like to become a data scientist, then you should work on these skills:
- Basic Probability and Statistics: Nothing too fancy, just the rudimentary stuff
- SQL: You are probably familiar with this (weird) language. You have probably used an ORM to interact with different databases. Learn a little bit more about it: window functions, CTEs, triggers, good SQL style guide, and so on
- Modeling: Again, nothing too fancy. Learn some good models to use and when to use them. Read the documentation and tutorials online when needed. This skill also requires domain knowledge about the things you are working at (ranges from health insurance to warehouse logistics)
- Data Visualization: Data analysis is not very valuable until you turn it into a graph: could be a map, a time series, a 3D pie chart (just kidding, don’t do that please), or anything else
- Reporting: Once you have some solid insights, you should make it available and organized into a compelling report. It could be a document or a dashboard (always prefer these)
- Communication: Finally, you have produced a report and/or a dashboard. You should be at ease when discussing it with your colleagues and superiors. This is a very hard skill to master but totally worth it in the long (and the short!) run
You’ll love the two stories I’ve mentioned below as well about how two software engineers successfully transitioned into a data scientist role. The articles detail the learning path these people took to achieve their dream:
- How I became a Data Scientist after 8 years working as a Software Test Engineer
- How I became a Data Science Analyst from a Software developer
2. Career Transition from Finance to Data Science – Richard Saldanha
Finance seems like a natural fit for data science, doesn’t it? It’s a numbers field and that blends in nicely with the data science space. It’s no coincidence that the BFSI sector is leading the way in data science adoption!
So if you’re coming from an accounts/finance background, you’re already halfway to achieving your dream of getting a data science role.
Answer by – Richard Saldanha
If you want to work as a data scientist in finance, you will probably need most (if not all) of the following attributes:
- A degree in mathematics/statistics, computer science, physics, engineering or subject with significant mathematical content
- An ability to program in multiple languages (both compiled and interpreted) such a C/C++, S (e.g. as implemented in R), Matlab, Python and/or Java
- Good database skills (i.e. at least SQL programming) in any classical RDBMS (for example, MySQL, PostgreSQL, Oracle, SQL Server)
- An adeptness with handling time-series data from Bloomberg, Reuters or any of the myriad financial data streams available
However, there are also two very important characteristics of people doing data science jobs in finance which are less frequently discussed:
- You’ll need to be able to communicate mathematical ideas well both verbally and visually to non-specialists
- You’ll need to know how to harness their mathematical training to solve genuine commercial problems
Alongside all this, you’ll need a good understanding of optimization (underpinned by solid linear algebra and calculus learned in school), of statistical inference, simulation, multivariate analysis, and proper data visualization.
I also recommend checking out this article which talks about the applications of data science in the finance industry.
3. Should a UX Designer/Researcher Become a Data Scientist? – Chris R. Becker
This is an intriguing career transition! I’ll be honest – I hadn’t considered a UX person wanting to transition into data science.
This answer by Chris R. Becker focuses on learning data science tools keeping UX experience in mind. He states some of the tools which UX designers are already using and how those tools could be used for data science purposes. He emphasizes more on working with a data science team to dive deeper into critical data science topics.
Answer by –
UX researchers are more often than not already using data (qualitative and quantitative) tools. Most are already using low hang data science. This includes Google Analytics, surveys, user polls, Excel, JSON, and user testing data (among other things).
These are data tools and methods essential to doing UX and finding patterns in the data. However, they are limited in scope and reach.
As a UX researcher, I would much rather work with a data scientist than have to learn a whole other profession to do my job effectively. I think it is crucial to know what is possible with data and then seek out that expertise as needed. Obviously fluency in data science tools will go along way in collaborating with a data scientist in this case. Seek out team members who can help you find new and relevant patterns in your research data.
I took this answer from an Analytics Vidhya article itself given how relevant it is. A lot of application developers want to transition into data science but are unsure if they are qualified enough for it.
This article by Ankita Ghoshal will put any doubts to rest! She explains, in thorough detail, how you can make this career transition successfully if you get started now!
Answer by – Ankita Ghoshal
The best way of penetrating into a new field is by first understanding the current technologies. The buzzwords back in 2016 were, as you might have guessed, ‘Data Science’ and ‘Machine Learning’.
I had vaguely heard about these terms through online articles. I started exploring career options in this field and found that Statistics was the base of Data Science. This meshed perfectly with my interests – Statistics has always fascinated me. There is nothing better than working in a field that you love!
A quick Google search on ‘Analytics Machine Learning Tutorials’ led me to India’s largest data science community, ‘Analytics Vidhya’. I went through their articles on educational institutions providing courses for careers in Data Science.
I spent most of my professional career in programming before switching to Data Science. As a kid, I studied statistics but those concepts were long forgotten (as I’m sure you’ll relate to!). Making this transition was indeed tough, but not impossible.
The key is to never stop learning. During this switch, I realized that you don’t need to unlearn your existing skills to pick up a new one. I used my programming skills as a bridge between IT and Data Science to structure my machine learning code more logically.
This transition also helped me understand that the presentation of project results varies significantly from industry to industry.
For example, in the IT industry, the output of a web development project is a web page that is completely understandable by the stakeholders. In the world of data science, the output is (usually) numbers. It is the role of a data science professional to reveal these numbers to the customers/stakeholders using an indicative story.
For those who are ready to start your transition into Data Science, I recommend reading the below suggestions carefully:
- Ask yourself this – are you are really interested in data science and are you a good fit? Don’t just fall for the glamour and hype. There are plenty of resources available online, such as the various articles and blogs on Analytics Vidhya, to get a feel of what this field is all about
- Statisticians and programming professionals will undoubtedly have a bit of an advantage. But here’s the good news – the transition is possible even for non-technical people. The primary thing that matters is your thought process and skill of questioning and analyzing the information at hand
- If you are an experienced professional in any other sector, get ready to be treated as a relative fresher when you make the switch to data science. It might be quite difficult for many people to accept a transition where you have to forego some years of seniority. I understand that. But if you give your best, then this industry will bestow even more knowledge, greater successes and awesome salary hikes on you
Note: You’ll also love the below story about how an IT person, after working in the field for a decade, transitioned into data science:
One of the most common questions we see is – can I become a data scientist without a technical/engineering background? The short answer – yes!
The below answer by Arun summarizes our thoughts perfectly. You don’t necessarily need a Ph.D. or even a programming background to start (though that might be helpful if you have that experience!).
Answer by – Arun Korupolu
There is no background required for you to become a data scientist in the long-run, it’s all about your interest and you asking whether you are interested to work with data and envisage yourself in a role where data and decision making are aligned.
I can suggest a high-level learning pathway, but individual learning needs with regards to time and effort might require appropriate tweaking in these steps.
Beginners ideally would need to start learning programming, in case they have no prior experience. You can follow this learning in three steps:
- Learn Programming (R or Python) and become proficient in that language
- Gain the knowledge in these subjects – Intermediate Statistics & Probability, College Algebra, Linear Algebra, Machine Learning algorithms and methods
- Work with independent projects. Try to implement your learning in a step by step fashion while solving the objectives of these projects
My suggestion is a high-level overview of what you can do to start, but you will find the best path once you begin learning by doing.
Note: You can also check out Analytics Vidhya’s comprehensive and free learning path to become a data scientist in 2020.
Are you looking to transition into data science from one of these backgrounds (or a completely different one)? Our community would love to hear your story! Stories and experiences like these help future career transitioners as well.
As I mentioned at the start of the article, you should definitely enroll in either of the two learning paths depending on your career aspirations: