Millions of transactions on a daily basis is a testament to the popularity of Amazon and the trust that customers place on the online retailer.
Naveed Ahmed Janvekar, our next pathbreaker, Senior Data Scientist at Amazon (Seattle), solves various fraud and abuse related problems for Amazon’s third-party marketplace.
Naveed talks to Shyam Krishnamurthy from The Interview Portal about relishing the challenge of providing a trustworthy shopping experience to millions of customers worldwide by identifying and analyzing questionable transactions among billions of data.
For students, always remember that data science should be guided by a larger vision to solve the most impactful business problems at scale.
Naveed, can you take us through your background?
I grew up in Bangalore, India. My father is a retired civil engineer and mother is a homemaker. I studied Electronics and Communications Engineering during my Bachelor’s and Information Sciences during my Masters. My initial interest was in problem solving using machines. In my free time I give a lot of importance to physical fitness – a healthy body is important for a healthy mind. My parents and my sisters have always been my pillars of strength for the success I have achieved today.
What did you do for graduation/post graduation?
In 2009, I enrolled for a Bachelor’s of Engineering in Electronics and Communication at B.M.S College of Engineering, Bangalore, India. After graduation, I joined Fidelity Investments as a Java Developer, where I was primarily working on developing software apps for efficient financial document management. While I was working at Fidelity I developed a passion for data due to engagement with a couple of financial analysts. In 2014, I got admission in a Master of Science in Information Sciences at The University of Texas at Dallas, USA. My areas of focus during my Master’s were Machine Learning, Advanced Business Intelligence, Big Data and Statistics.
What were some of the influences that led you to a career in Data Science?
My interest in the Machine Learning space grew due to a couple of data analytics tasks I did in my personal time for expense management for a small population. Doing so, I was able to extract insights and trends from data and could also establish patterns in people’s spending. After this, I started spending more time researching Machine Learning and how we could leverage it to model repetitive patterns to predict future outcomes and use it to our advantage to solve critical problems at scale. This is when I decided to pursue my Master’s in Information Science with a specialization in Machine Learning and Analytics. In my career journey to become a Leading Data Scientist, I also took on various analytical roles such as Business Analyst, Business Intelligence Developer at Ad tech startups, Accounting firms and ECommerce giants. Certain key influencers in my life have been renowned Data Scientists such as Andrew NG and DJ Patil.
Tell us about your career path
To pursue a career as a Data Scientist I had to first get all the advanced knowledge and skills for this domain, which I achieved through my Master’s at The University of Texas at Dallas. I enrolled in subjects that were focused in Data Science such as Machine Learning, Advanced Business Intelligence, Big Data, Data Visualization, Data Warehousing, Statistics. This helped me set a strong foundation in the field of Data Science. At The University of Texas, I also received the Dean’s Excellence Scholarship as well as the award of Scholar of High Distinction, which helped me with tuition fees and recognition.
I got Dean’s Excellence Scholarship as a result of excelling in all my enrolled courses in 1st year of my Master’s. I maintained a 4.0 GPA (perfect score). As a result of receiving this scholarship I was inducted into Beta Gamma Sigma society which is an international Business Honor Society. They recognize and honor top performing students from around the world. Notable members include Nobel Prize winners, Olympians, Inventors, CEOs of major global companies and nonprofit organizations, Deans of the top business schools, and others who are making the world a better place through social enterprise, service, and leadership.
In order to get practical exposure in the corporate world I joined an Ad Tech startup called Nanigans in Boston, where I worked on analyzing ways to optimize ad spend of customers . After completion of my Master’s, I joined KPMG as a Business Intelligence Developer where I worked on building various data analytics apps for the company.
In the year 2017, I got an opportunity to work at Amazon in Seattle as a Business Analyst to work on important customer facing problems. As a part of the role, I started applying various machine learning algorithms to solve critical problems and dwelled into research to better serve my company and customers. All of my work translated into better customer experience and positive business impact and fast tracked my progress to becoming a Lead Data Scientist.
How did you get your first break?
I would say my internship at Nanigans in Boston was my first break. What helped me get this internship was my coursework at the university and my experience in using web analytics to improve traffic on the website of an “Ethnic Wear” ECommerce startup. Additionally, I would encourage Data Science aspirants to participate in as many Machine Learning competitions as possible, write research papers, collaborate with senior scientists in the field. This goes a long way in making connections in the domain who can then help with the right opportunities.
What were some of the challenges you faced? How did you address them?
Challenge 1: Finding the right problems to solve. Oftentimes we are given a bunch of problems to solve not knowing the actual impact that a problem has on customers. It is important to seize the opportunity of solving a problem so you know that your efforts in building a ML solution will drive significant impact. In order to navigate this challenge I typically have meetings with cross functional stakeholders, do opportunity sizing analysis before jumping into an ML solution.
Challenge 2: Finding good mentors in the domain. It is very important to have a good mentor at the very beginning at least to guide you on the skills that you need to develop to be a successful Data Scientist. Not all people you reach out to might have the bandwidth to mentor you, but if you can draft a story around your current profile, skill set, aspirations, and how a mentor’s work can benefit you and then reach out to senior scientists, they would be more than willing to engage rather than if it is just a cold ask request.
Where do you work now? What problems do you solve in Data Science?
I am currently working as a Lead Data Scientist at Amazon in Seattle. I solve various Fraud and Abuse related problems for Amazon’s third-party marketplace. Some key skills that are important for my job are Machine Learning knowledge, Python programming, Spark, Statistics, SQL, Data Visualization.
Some of the use cases I work on involves identifying Sellers or Buyers on the platform that are violating Amazon’s policies. This involves using huge volume of data (e.g. billions of records) to make high confidence predictions. Additionally, as a Data Scientist one needs to make sure that the model performance is at optimal levels through new features, new algorithms, by training the model with the latest data and so on.
Being a Lead Data Scientist I spend significant time in building out the Data Science vision for Amazon and have many meetings with various cross functional team members. When I am not in meetings, I work on developing novel methodologies for Amazon using Machine Learning.
The aspect that I love most about my job is that every day there is a new challenge waiting to be solved that impacts millions of customers worldwide.
How does your work benefit society?
By providing a trustworthy shopping experience to millions of customers worldwide.
Tell us an example of a specific memorable work you did that is very close to you!
A specific memorable moment for me was when customers around the world acknowledged that a specific problem they were facing was no longer an issue to them due to my work and me being acknowledged as one of the top innovators at my company.
Your advice to students based on your experience?
For students pursuing a career in Data Science, understand the nitty gritty of Machine Learning algorithms, the math behind it. Focus on business problems that need to be solved and use what is the best approach to solve the problem and not necessarily the most sophisticated one. Be good at object oriented programming such as Python and Data Extraction and Manipulation using SQL, Hive programming. Collaborate with senior scientists in the field from early on in your career and participate in research papers.
I would like to invent more novel Machine Learning methodologies for my company to solve customer facing problems in the US and worldwide as well as research more aspects of improvement in Machine Learning models using Active Learning
In the coming years, I also plan to have a Data Science Online Edu platform.