The demand for data scientists is up, even as a 2017 IBM study titled ‘How the demand for data science skills is disrupting the job market’ projected a 28% spike in Data Scientist and Data Analyst roles by 2020. Newer industries too are roping in data scientists to craft their business solutions. Findings indicate that India offers multiple opportunities for data analysis enthusiasts to thrive, even with basic skills. If you’re new to data science, keep in mind that this field is impacting practically every industry; for instance, medical image analysis in healthcare and credit risk modelling for commercial loans in BFSI hinges on data science.
Moreover, the applications of data science aren’t all that niche. Think YouTube. Anyone who uses the platform encounters data science at work through machine learning and deep neural networks at play in the video recommendation algorithm. What YouTube does is match videos to your taste – and that means analysing tons of data points including time spent on video, video popularity rate, channels you have watched in the past, what you avoid watching and so on. The result is thousands or millions of versions of YouTube around the globe, all maximising the platform’s capability to keep you hooked on and earn through ads. In short, data is being looked on as the new ‘oil’ of the economy and, in that respect, data science is in vogue.
However, you may be wondering what the career of a data scientist has to do with that of an automation tester. Well, there are a couple of considerations to begin with:
- Data scientists are becoming popular whereas a Harvey Nash study noted that ~70% of testers feared being automated out of a job by ~2026. Data scientists do not have similar fears, at least as of now. Job security is surely better for an automation tester than a manual tester, but the prospects that a career in data science afford are worth exploring.
- Recent evaluations of a data scientist’s career suggest that there are multiple entry points to the field. In essence, if you have the technical skills required, you can be a data scientist.
This brings up the next and most important question. Can an automation tester become a data scientist? For a clear-cut answer, read on.
Yes, you can, and you have an upper hand – ask Microsoft’s Ken Johnston!
The proof is in the pudding and Ken Johnston’s career progression leaves no doubt to the fact that it is possible to jump from testing to data science. Search for similar career trajectories online and you’ll get a bunch of stories! However, when a Test Manager turned Director of Test Excellence at Microsoft goes on to become a Principal Data Science and Data Engineering Manager at the tech giant, the evidence is compelling.
For Ken Johnston, the switch involved developments such as cloud, Agile, and, importantly, the concept of minimum viable quality (MVQ). Rather than a specific path, what makes the journey from testing to data science possible, even easy, is that software testing is innately data-driven. Ken Johnston himself notes that:
“Good testers have an innate touch for data that makes them great candidates to become data scientists.”
Here’s why becoming a data scientist should be easy for you as an automation tester.
- Testers deal with sets of data and know how to use it optimally
- Testers know how to look at data with a sense of critique
- Testing and data science are both empirical fields seeking answers
- Both fields are mainly concerned with data and not the code
- The switch to data science involves a shift in perspective – scientifically and psychologically
- Testers already deal with terabyte-large complex systems
Your expertise in Python, R, and SQL gives you a head start as a data scientist
So, it’s established: moving to data science from automation testing is not so much a jump as it is a transition. But wait! The commonality between the two lies not just in the conceptual level – that both fields involve getting your hands dirty with loads of data – but there is a reasonable amount of technical skill overlap too. This makes it easier for you to become an accomplished data scientist engineer.
As per a compilation on Towards Data Science, the top data scientist skills in 2020 include:
- Probability and statistics
- Linear algebra and multivariate calculus
- Python, R, SQL, Java, MATLAB
- Data Wrangling
- Database management
- Data visualisation
- Machine learning
- MS Excel
If you’re wondering how to become a data scientist, the answer is simple: master these skills. As an automation tester, the path would involve starting at areas where there is already some overlap. Some skills or domains you may already be fairly conversant in are:
- Python, R, SQL
- Machine learning basics
- Cloud computing
- MS Excel
- Probability and statistics
So, you’re not starting from scratch. Acquiring the required technical skills could be a matter of:
- Brushing up on your math
- Learning advanced statistics
- Digging deeper into Python, R, and SQL and learning tools like Tableau and Julia
- Going deep into deep learning/ machine learning
- Learning techniques for data wrangling
- Exploring data science concepts
The final word: No matter the developments in data science right now, being an automation tester means that you have already made forays into the domain of data science. Now, it’s a matter of shifting gears and tweaking the engine for your new route.
You have plenty of readily-available resources to build your data scientist tool kit
Getting practical, your next step is probably going to be one or a combination of these three: pick up leading data science books, take online courses, or go to a data science institute.
In terms of books, you can start off with titles like:
- Data Science from Scratch by Joel Grus
- R for Data Science by Garrett Grolemund and Hadley Wickham
- Python for Data Analysis: Data Wrangling with Pandas, NumPy, and IPython by Wes McKinney
- Storytelling with Data: A Data Visualization Guide for Business Professionals by Cole Nussbaumer Knaflic
- Deep Learning by Aaron Courville, Ian Goodfellow, and Yoshua Bengio
Since self-study from books often lacks structure, you may want to take an online course as well. Make sure to pick the best certificate course you can find. You can enrol for these from the best institutes, filter for course durations and skills, and, best of all, learn without leaving your current job.
Some course curricula you can peruse through are those of:
- Python for Data Science and Machine Learning Bootcamp (Jose Portilla, Udemy)
- MicroMasters Program in Statistics and Data Science (MITx, EdX)
- IBM Data Science (IBM, Coursera)
Likewise, if you’re looking for a 6-to 24-month full-time course that would offer you better lab experience, you could consider courses such as:
- PGP in Data Science (INSOFE)
- MBA in Data Sciences and Data Analytics (Symbiosis)
- M. Tech in Data Science (IIT Hyderabad)
Automation testers work on hands-on projects and that’s key to data science too
Both fields, automation testing and data science, require you to possess an analytical mind. This skill helps testers unearth bugs in the system under test, and enables data scientists to glean meaningful insights, such as customer behaviour, from the data at hand. But, for analytical thinking to develop, one must have real-world experience – and there is no dearth of hands-on experience for the automation tester!
The whole process of transforming ‘raw’ data into a processable format and then analysing it critically, scientifically – even psychologically – can be a challenge, but as an automation tester, a long haul is nothing out of the ordinary. Your task is to gain relevant experience. This is crucial, especially since you are moving laterally within the job market. In other words, the probability of a top organisation hiring you improves the moment you have data science experience, and not just skills, on your resume.
Two methods help:
- Take a professional course where there are lab sessions
- Look for a post within your existing organisation where you can work on data science projects
Then, with skills and experience backing your profile, sign up with Talent500. This is a great way to be found by Fortune 500 companies seeking the top 10% of talent. In fact, Talent500’s dynamic skill assessment and ML algorithms will match your skills to relevant data scientist posts. This way you’ll get discovered even as you set out to discover more from data!