Springboard Mentor Spotlight: Rahul Sagrolikar

Meet Rahul Sagrolikar, a Data Science Lead at Amazon and mentor for the Springboard Data Science Career Track.

As someone who helps students from non-technical backgrounds prepare for careers in data science, Rahul Sagrolikar is somewhat of a career switcher himself. After pursuing an undergraduate degree in mechanical engineering, Sagrolikar went back to school for an MBA with a focus on mathematics and statistics. Since then he’s held data science roles in major companies, including Vodafone and Thomas Cook.

Sagrolikar says that when he first entered the profession, data science was still such a new field that he routinely found himself as the sole data professional on his team, so receiving mentorship was out of the question, and he had to vouch hard for his work. Now that he’s a mentor, Sagrolikar says he’s happy to be able to provide students with a key resource that he never had.

This conversation has been lightly edited for brevity and clarity.

What do you do at Amazon?

I work at Amazon within the marketing team for Prime Video. My job is to use data science techniques to optimize the different marketing channels we use to target customers.

How long have you been a Springboard mentor?

Almost three years.

What do you like most about being a mentor at Springboard?

I learn from my students every day. Sometimes there are industry trends I’m not aware of because I’m busy with work commitments, and students will ask them about. So then I have to go on the internet and do my research. In this way, they help me stay up-to-date. I love that there is that give and take.

When I started my journey in data science, I never had a mentor who could guide me. I learned on the job by being the only data scientist on the team. But the Springboard platform gives students the opportunity to be better prepared when they join any industry because they will know exactly what is expected of them and how they need to perform on a daily basis. Being a part of it, I am giving them [the mentorship] which I never got, which is exciting. I am shaping their career, helping them reach their potential, and that is the most satisfying thing for me since I joined Springboard.

What is something you have learned from your mentees?

One of the most important things is presentation skills. Coming from a technical background, I always had more interest in coding and problem-solving and other technical skills, but I was never good at presentations, PowerPoints, and writing reports.

I’m often inspired by how my students structure and format their PowerPoint presentations and reports when they complete their capstone projects. Sometimes I’ll even recreate those templates and use them for my own presentations.

Do your students know that they’ve inspired you?

Absolutely, I always tell them. Many of these students are blessed with good presentation skills—probably because they are just starting their journey in data science and have come from another field—so I just help them polish their technical skills. I tell them how advantageous it is for them to be good at both. It gives me immense satisfaction knowing that I’m a part of their journey and that I can take cues from them to use in my day-to-day work.

How has being a mentor helped your own career development?

Because I am managing so many students, it has helped me become more accountable in terms of time management. I need to take care of my personal priorities, my work priorities, and then the additional responsibilities of shaping students’ careers as well. As you grow in your company, you start managing more and more people and time management becomes a critical skill, so I feel that this particular aspect [of mentoring] has helped me in my career development.

What is the most interesting project you’ve worked on in your career?

A few years ago I was the head of data science at Thomas Cook, a travel company in the UK that provides holiday packages to customers. Before the company went into bankruptcy a year ago, it was a renowned British institution that had been around for 150 years. Incidentally, Thomas Cook was the first person who thought of this tourism concept of organizing end-to-end travel experiences.

We were tasked with a project to improve our hotel listings. Under the current system, we weren’t converting many customers. So we looked into all the historical data we had on our customers, such as which holidays they book and which locations they prefer. Then we built a real-time recommendation engine. Based on the user’s journey with Thomas Cook, from the moment they entered our website, the model would immediately know what holidays to recommend to that individual based on historical web profiling and site interactions. Finally, we did an A/B test to see how well the model worked compared with the original recommendations we had for our clients. We saw an improvement of around 3-4 percent with the new recommendation engine in terms of the number of users booking vacations through our site.

What advice would you give to someone who is just starting out as a Springboard mentor?

Make sure you empathize with your students and be patient with them. Sometimes it gets frustrating if a mentee doesn’t understand a concept. They may come from a non-statistical background. Come up with simple examples to relate complex data science concepts and emphasize the important points they need to take into consideration. The course can feel overwhelming. You need to break the course down into simple steps for them. Let them know what concepts they absolutely must learn and which ones they can afford to skip. You need to create short-term and long-term goals and provide accountability for them so they stay motivated in the course.

What do you mean by accountability?

Your job is to simplify whatever analysis or concept they are learning so they can feel more confident about it. You can provide them with examples of how that concept is used in the industry.

What happens sometimes is that you have a lot of mathematical equations or derivations on the backend when you are trying to learn an algorithm and it gets overwhelming. In that scenario, you can advise them to set aside those equations and focus first on the utilization part—how, when, and why you should use this algorithm. Once the student understands that part, you can prompt them to go back and learn about the mathematical equations behind it.

Have you ever seen a Capstone project and thought, ‘Wow, I wish I thought of this!’ or ‘We should implement this at my company?‘

Yes, I remember one project by a student who worked for a company that specializes in 3D printing. You provide the specifications to the robot and it creates a 3D model for you. He told me that sometimes the quality of the print was good and sometimes it wasn’t, but they weren’t sure what parameters they needed to give the machine to get a consistently good print.

So we converted this into a data science problem. We collected historical data from his company on what specifications or parameters were entered into the machine and what the end result looked like—was it a good, moderate, or bad quality print? We accumulated the data to understand the optimum parameter values we need to feed into the machine for the best end result. It is a classical example of using data science to solve a day-to-day problem.

What is the most interesting dataset a student has used?

Typically, students will use datasets from Kaggle for their capstone project, but the student I mentioned above took the extra effort of getting data from his company. It was inspiring for him to learn that when you are dealing with a situation where you need to optimize multiple parameters to get the best output, you can always convert it into a data science problem.

What soft skills have helped you succeed in your data science career?

Business acumen—that’s the most important soft skill you need to be a successful data scientist. You can learn about Python packages and how to implement algorithms on a dataset. But business acumen helps you understand how to structure the dataset. It all depends on how you create the right dataset for your model to work. If you fail at that, no matter how much data you analyze or however complex an algorithm you use, you won’t get good results.

When a business comes to you with a problem, you need to understand how the business functions, which KPIs the business relies on, and whether those KPIs are present in the datasets you are working on.

The other skill I would recommend is the art of storytelling. You know your technical concepts well, but you have to translate them for a non-technical audience—mainly your leadership team. They want to understand why you’re implementing a certain algorithm and how it benefits the company, so you need to be able to tell a good story of whatever analyses you’ve done.

If you fail to convince the leadership team, no matter how good or complex of a model you build, you are not going to get buy-in or sign-off for your projects.

Any closing thoughts you’d like to share?

I would recommend that anyone who has the capacity to mentor should join Springboard because you are helping to guide the future generation. Who knows, maybe tomorrow they will become a better data scientist than you are. We need to capitalize on that because the world needs more data scientists. So when you reach a higher level in the company and you need to hire a talented data scientist, you can pick from this talent pool, so it helps you in that way as well. That is why I would recommend to all of those who have 5-10 years of experience in this field to actively engage as a Springboard mentor.

Want to become a data science mentor? Apply here!

Since you’re here…
Curious about a career in data science? Experiment with our free data science learning path, or join our Data Science Bootcamp, where you’ll get your tuition back if you don’t land a job after graduating. We’re confident because our courses work – check out our student success stories to get inspired.