Shweta Gupta, VP – Technology at Digital Vidya: Only techies know the awful feeling when you can’t talk about the nuances of your work with general people. And that for me is a struggle! Being a hardcore techie, someone who has been a technologist for nearly two decades, I am really passionate about talking about my field of work. But generally, I don’t get such opportunities out of my workplace. Obviously, this is something I do not appreciate much. But now, with Digital Vidya’s Industry Insights Series on a roll, I am more than excited to get a chance to talk to the leaders of the Indian Data Science Space.
Whenever I have gone on to interview these exciting people, there is a thing or two that I have learnt. And if I talk about my fascinations, then the kind of work each of these experts is doing are new horizons in themselves. Talking to Akshay Sehgal, was an experience. The kind of work he is doing is a discovery. Plus, he was willing to talk about the subject at length which acts as an eye-opener. Now, I think I have spoken enough. You are more interested in reading from Akshay than to learn my thoughts. Let’s dive into the details of the interview right away.
After completing his Computer Engineering from Army Institute of Technology, Pune, Akshay Sehgal was introduced to Predictive Analytics & Data Science at Mu-Sigma where he was able to develop skills in Machine Learning, Image Processing, Natural Language Processing and Computational Mathematics. After 3+ years of working with multiple US/EU based clients on complex problems from retail, banking, telecom and casino industry, Akshay collaborated to set-up a Data Science product-based startup called iPredictt Labs where he handled the products & research. For the next 2-3 years, Akshay built intelligent, data-powered products in the Advertising and HR domain. Following iPredictt, he joined Reliance Industries as a Lead Data Scientist where he now applies his expertise building the new age, enterprise-scale data science solutions across multiple domains from employee services to finance.
How did you get into Data Analytics? What interested you in learning Data Analytics?
Akshay Sehgal: I think my interest towards logic & mathematics, along with my computer science background introduced me naturally to data analytics. More than anything else, what kept me passionate about the subject is the sheer volume of learning needed to keep up to the latest standards. For people like me who tend to get bored easily with monotonous work, learning and applying new algorithms every second day is enough to keep me hooked to data analytics.
What was the first dataset you remember working with? What did you do with it?
Akshay Sehgal: My first dataset was video data from a US-based superstore which was aisle camera footage of people walking around the store making purchases. Our objective was to analyze the footage to track customer movement around the floors. We built heatmaps to show areas with high dwell time indicating specific shelves where customers spent relatively higher time, allowing the company to make better decisions on how they merchandized their products inside the store.
Was there a specific “aha” moment when you realized the power of data?
Akshay Sehgal: There have been so many such moments and honestly that is what makes data science so addictive. But if I were to pinpoint a specific moment, it was the day I applied Tomas Mikolov’s Word2Vec algorithm on text data. Just the fact of how different words are placed around other words lets you encode some information about their meaning. It was almost magic! My algorithm could understand the meaning of a piece of text without anything but a large amount of related text to train it on!
What is your typical day-in-a-life in your current job? Where do you spend most of your time?
Akshay Sehgal: Brainstorming and implementing white papers is what a larger part of my day is spent on. The problems that I work on usually don’t have any ready-made solutions in the market and so it becomes necessary to design them from scratch by playing with different models and methods. A decent chunk of my time is also spent on working with product teams, functional experts, dev-ops teams and UX experts to scope the business objectives and product design at hand.
How do you stay updated on the latest trends in Data Analytics? Which are the Data Analytics resources (i.e. blogs/websites/apps) you visit regularly?
Akshay Sehgal: Constantly reading whitepapers and finding online material on applied mathematics helps me stay updated. I have also joined many data science/analytics-based groups on social media, which provides me with a lot of fresh material to read on daily basis.
Share the names of 3 people that you follow in the field of Data Science.
Akshay Sehgal: Andrew Ng, Kenneth Cukier, Elon Musk, the list can go on and on…
Team, Skills and Tools
Which are your favourite Data Analytics Tools that you use to perform your job, and what are the other tools used widely in your team?
Akshay Sehgal: Python primarily, with a bit of R, Octave and Matlab.
What are the different roles and skills within your data team?
Akshay Sehgal: My team comprises of senior and junior data scientists who are individually experts at one or more phases of the data science process cycle such as algorithm mathematics, code optimization or python deployment.
Help describe some examples of the kind of problems your team is solving in this year?
Akshay Sehgal: Couldn’t reveal too much, but we are working on some geo-spatial route matching algorithms, document matching, anomaly detection and a very interesting problem on expression & empathy detection.
How do you measure the performance of your team?
Akshay Sehgal: Passion and focus towards work is a BIG part of what we do here, and it shows in the form of innovativeness and project ownership. This usually remains my primary criteria for measuring performance. Quality of deliverables over quantity!
Industry Readiness for Data Science
Are the industries looking to understand what they can do with data? Do they have the required data in place?
Akshay Sehgal: I think almost every industry is looking to utilize data science as a core part of their processes. This is due to the fact that all industries are now a positive sum game. Meaning that there is no fixed sized pie, and innovation can help increase the size of the pie for them. This makes leveraging intelligent data powered algorithms for improving quality and quality of existing process, quite appealing. The biggest problem I feel is that a large number of companies don’t really understand what data science can help them with, and since there is already a massive demand, companies end up wanting data science but not knowing how to leverage it properly. Another problem is the gap in understanding the importance of data. The data science algorithms are still garbage in garbage out. If you don’t have good data, no number of intelligent algorithms can help you!
Advice to Aspiring Data Scientists
According to you, what are the top skills, both technical and soft-skills that are needed for Data Analysts and Data Scientists?
Akshay Sehgal: It’s necessary to be able to play with data and apply algorithms so knowledge of at least one programming tool is important (Python, R, SAS, Octave, Matlab). The biggest missing skillset in most aspiring data scientists is mathematics. While most are well versed with statistics and probability, they lack basics in linear algebra, calculus and topology which are critical for understanding today’s algorithms in depth. From a soft skill perspective, the only skill I think a data scientist requires is the ability to brainstorm.
How much focus should aspiring data practitioners do in working with messy, noisy data? What are the other areas that they must build their expertise in?
Akshay Sehgal: A lot! The reality is, that we will never get clean data and so it becomes almost necessary for a data scientist to make it a part of their expertise. The fun bit is that there are some interesting algorithms which can be applied to trivial stuff such as missing data points, which can help keeping you engaged during the process of handling messy data.
Secondly, it’s important to build capability to understanding at least the intuition of the mathematics behind the algorithms and models you apply.
What is your advice for newbies, Data Science students or practitioners who are looking at building a career in Data Analytics industry?
Akshay Sehgal:
- Practice using the tools constantly to build intuition on how different data types behave.
- Try being agnostic to tools and learn the underlying methodology behind machine learning
- Focus on linear algebra as much as statistics. It’s more widely used in data science than you can imagine.
- Compete on data hackathons such as Kaggle and Analytics Vidhya. Read and learn from what other participants are doing.
- Learn the art of reading complex whitepapers. It comes with practice and smart studying.
What are the changing trends that you foresee in the field of Data Science and what do you recommend the current crop of data analysts do to keep pace?
Akshay Sehgal: Constant learning is necessary for this field. There are too many hubs of innovation around the globe working 24X7 to create more and more advanced methods of solving different data problems. You need to keep up with the progress else you will lose out.
Would you like to share few words about the work we are doing at Digital Vidya in developing Data Analytics Talent for the industry?
Akshay Sehgal:
Digital Vidya is doing a great job at bringing data analytics, the field that I am so passionate about, to the rest of the world! There are tons of learning material across the web, but too little guidance on how to consume it. Digital Vidya handles this need quite impressively.
To know more about Akshay Sehgal, you can check out his LinkedIn.
Are you inspired by the opportunity of Data Science? Start your journey by attending our upcoming orientation session on Data Science for Career & Business Growth. It’s online and Free :).
Content was very useful and informative. Thank you for sharing.