Presenting to you the INSAID Spotlight Budding Data Science Leader interview series. This is a series of interviews of budding data science leaders, enrolled with INSAID in different courses. These students coming from diverse backgrounds and even different fields, have rich experience in their domains. They have interesting views to share with the world, their experience in the industry, what brought them to the field of data science and many other such interesting aspects. These interviews will enrich the readers about the insights, trends and many other related points.
In a recent conversation, we spoke to Lava Kumar who is enrolled in the GCDAI program at INSAID.
Name: Lava Kumar
Current Organisation: Appfire
Total Experience: 10 years
Batch: Global Certificate in Data Science and AI (GCDAI) November 2018
Ankita: My first question to you would be about your career journey. What has been your professional and educational qualifications and your experience?
Lava: I started my career in 2008. I had a campus placement in Jawaharlal Nehru Technical University and got selected in Wipro. There I worked as a Software Developer for close to six years.
During my tenure, I work with various clients like Microsoft and CA Technology, where I had a chance to explore a variety of technologies. I mostly got interested in Java as it was my core strength.
I then moved on to a company called enVista, which is into the retail industry and here I got a very good exposure to working in the product environment. I was mainly working on the UI development, and also got a chance to write a couple of APIs using Java on the backend.
This really exposed me to good software development environment, where I spent close to three years working on the product. It was a full stack development role where I got involved in building product architectures and taking QA decisions in terms of how to scale the product and how to take different decisions.
After that, I moved on to a company called Appfire and here we have a very good client base, in terms of serving a marketplace. So, we have close to 1000 installations for our add-ons and we simply build apps on top of tools like Confluence In Jira, which are mainly used by major companies. This was the place where I got a very good exposure in terms of handling huge amount of data.
In this company, we did a lot of data analytics. I would also want to reiterate that Python has been my strength and have been using machine learning tools to understand data. In this attempt, Tableau has been really good to visualize the data in the report.
The journey has been so far so good and my experience in working with data has enabled me to understand the data even better. This is all about how I explored the machine learning tools and algorithms. Finally, I joined INSAID to learn more about data science and related technologies.
Ankita: Good to hear that! Should I take it as your earlier understanding of Python made you enter the field of Data Science and Machine Learning?
Lava: To be honest, I had an affinity for mathematics since my school days. So, I really wanted to apply math in my areas of programming. I did that but it became interesting after I learned machine learning, where we use a lot of mathematical concepts and algorithms. That’s how I got into machine learning around 3 years ago.
Ankita: Okay! My next question to you is- what all tools and packages in Data Science and Machine Learning have you mastered?
Lava: I have explored data analysis, and have used linear and logistic regression. I have also used TensorFlow because it has great libraries and packages. It lets me do anything from scratch. I enjoy writing code from scratch. I’m into project details and being in this field, I would prefer using TensorFlow.
Ankita: When you entered the field of Data Science and Machine Learning, what challenges did you face? If there were any, how could you overcome them?
Lava: The major problem that I faced in one of my companies was that I couldn’t find a lot of data because the data they needed was not in one place. It was to be gathered from different sources. Data collection was my major problem and I solved it by writing basic packages.
I used Python to write packages to extract information from data sources. But the approach you adopt completely depends on the type of problem you’re solving.
In my opinion, if you’re stuck at something, it is better to approach people and see how they have solved the problem you’re facing. You can always approach the mentors in your field and talk to them, what it is that you’re trying to solve.
I solved this problem by going through different blogs and trying to figure out how did they approach this problem.
Apart from this, I also faced a problem as to which algorithm to go for; which algorithm would give better performance and accuracy.
When I tried approaching people with this problem, the feedback that I mostly got was to try every other algorithm. I was advised to compare the results and see which one is performing better and eventually choose the better one.
Yeah, I think that was the key; we cannot really understand which one is better at first sight, but then we have to try every other algorithm and note down their performance and working. For me, trying out different components worked for me.
Ankita: What according to you are the current trends in Data Science that you are most excited about?
Lava: The current trend that I’m most excited about is in the area of reinforcement learning, wherein the system will learn something on its own without the Data Scientists going in and monitoring it.
The system calculates on its own whether it is going in the right direction or not. If the result is positive, the answers are in the right direction but if the result is negative, the system knows that it has to implement some corrective actions to take the right decision. So I think there is a very good research going on in reinforcement learning.
There is yet another area- meta-learning. In this, the machine doesn’t need huge amounts of data to train on. Better learning will require a little amount of data as compared to other neural networks, where lots of data is required.
So I think there is a long way to go for reinforcement learning and meta-learning.
Ankita: Are there any blogs that you read to stay updated with the latest trends and happenings?
Lava: Towardsdatascience and machinelearningmastery are the blogs that I read and follow regularly. Apart from this, I follow a lot of podcast series. In addition to this, I read software engineering daily, machine learning by O’Reilly, Machine Learning 101 and Machine Learning by OCDevel. I listen to the podcasts and read the blogs and also correlate them with my current knowledge.
Ankita: If I talk about AI and Data Science influencers, are there any specific ones that you follow?
Lava: There are a lot of Data Scientists whom I follow. Andrew Ng and Dr. Kirk Borne are two names I would like to mention. Apart from this, I also follow many of them on Twitter and Facebook.
Ankita: My next question to you would be- at INSAID, students are encouraged to maintain high-quality GitHub profiles. Have you also built a GitHub profile? How do you think this will help you?
Lava: Yeah, definitely. I do a lot of recruiting in my company as well. So the first thing that I do is search for the candidate’s profile and check whether he has any hands-on experience.
As a matter of fact, most of the people contribute to open-source technology and very few ones work on the closed source and might not put their work on GitHub.
I would like to mention here how I got into the current company. It was based on my GitHub profile, which had close to 2000 contributions in a year. So that really helped the recruiters to zero in on me. The recruiter actually mentioned that he chose my profile over other 20, based on my GitHub profile.
So that was really positive for me because I contribute a lot.
I believe that maintaining a healthy GitHub profile and regularly contributing to the open-source technology will really give you an upper hand when it comes to being selected and also make it easy to select candidates when you are on the hiring front.
Ankita: That’s actually so good to hear. Your’s is one example of why one should maintain a healthy GitHub profile. Lava, my next question to you is- crafting a great data science resume is a critical part of getting shortlisted for Data Science roles. Tell us some ways in which you have improved your resume as a part of Data Science Career Launchpad.
Lava: Yes, I have done that and submitted it to the team on the Career Launchpad portal.
In the Career Launchpad session, we were told what is the importance of creating a concise resume and putting in it everything that will showcase your work and experience, in a precise way. As a result, my resume of eight pages is now of a single one.
This will definitely help me as no recruiter would read such a long resume. There is a notion that recruiters will spend an average of 6 seconds on a profile. So, I really took that seriously and made it of a single page, focusing on the major areas of my work. This will really help recruiters to focus on the major things that I’ve done.
Ankita: INSAID’s mission is to groom Data Leaders of tomorrow; what do you understand by Data Leaders? And how are they different from Data Scientists?
Lava: Data Leader is someone who will take the goal of the business forward; will build a team of data scientists from scratch; build a decoded data science ecosystem in the company.
Data scientists are the ones who work in democratizing the model and extending the results to the stakeholders. This, according to me, is the core difference between a data leader and data scientist.
Ankita: I would like to understand that because you are a working professional, how do you find time to study? Do you have a schedule? How are you able to maintain a balance between your work and studies?
Lava: I collaborate with my colleague Manish every day in the morning from 4:00 am to 6:00 am on data science and machine learning topics.
We just want to put consistent efforts into the stream so that we realize our aim of making it big in the field. It has been close to six months that we are following this routine.
It is really helpful. If you’re studying alone, you might not be able to track the time. But when you’re studying in a group, then you will exchange a lot of ideas.
I see a lot of advantage when it comes to studying together. I also study with my colleagues through the Zoom meeting app. We begin by simply logging in and then discuss the different techniques and tools we all have learned. So far, this process to study has really given a good understanding of the topic.
Ankita: I understand that is a good thing to do. So, now my question to you will be- what will be your advice to anyone who wants to start a career in data science and who is actually a fresher in this field?
Lava: My advice would be- to start in this field, you should follow a top-down approach.
For example, to learn any algorithm, try to implement the algorithms first and then try to understand what it is doing. This is what I suggest to freshers. They can really build something first and then try to analyze how things are working under the hood.
Also, joining a formal course like INSAID will take them in the right direction because they might have all the skills that are needed to succeed but the end ingredient to make you successful is the right guidance.
I would want to add that INSAID has given us a good direction and taught how Tableau is an extremely helpful tool for us. After undergoing the training at INSAID, I came to know that Tableau is such a helpful software, which otherwise I would not have known.
The content at INSAID is structured in a way that both freshers and experienced can benefit from it in a systematic way.
Ankita: Good! My last question- any review or feedback you have for the team here at INSAID; anything that you want to share regarding your journey at INSAID.
Lava: INSAID team has really done a wonderful job, in terms of taking the classes and Suchit is one of the many great faculty members who coached and trained us. His sessions were really interesting.
Even Manav and his sessions, particularly the Career Launchpad sessions, were very good. Students stand to gain a lot from these sessions. I think the team at INSAID is really on their toes when it comes to delivering and they are really doing a wonderful job in the data science area.
I think INSAID will soon stand out from the other players in the market.
Ankita: Thank you so much for such generous and genuine words. We’ll surely try our best to maintain the standards and live up to our promise of delivering the best education and training in the field of data science and artificial intelligence. Thank you so much for your time. All the best!
Lava: Thanks for your time, Ankita. Bye!