Interview with Dr. Pradeep K Mavuluri
Dr. Pradeep K Mavuluri
Lead Data Scientist, Machine Learning Consultant
Innovator, Speaker, Automation Specialist
Bengaluru, Karnataka, India
Q1. Please share your educational and professional journey?
I have done Ph.D. in Financial Econometrics from Hyderabad Central University in the area of Individual Stock Futures with applications of advanced time series models. Also, I have published one of the earlier publications on NSE High Frequency data.
Coming to professional journey, being from academic background, its Symphony’s IRI R&D division which attracted towards my corporate journey after my Ph.D. thesis submission, later, moved to product startups around data science applications. Currently, working as Master Data Scientist at HPE (a legacy organization, who are yet to adapt data driven insights decision making in this competitive world).
Feel immense proud to say that worked across domains namely., CPG/Retail, SCM (Supply Chain Management), IIoT (Industrial Internet of Things), and rare HR Analytics/MELT (Media, Entertainment, Leisure and Travel) domains for more than 14+ years.
Was also, Founder of start TATATVAI and guided more than 50+ juniors across corporate experience and several lectures and series at 10+ academic institutions.
Q2. How do you handle stress in this fast and competitive life?
Sticking to process (it can be with respect to business, quality or work you do). For instance, most of the time in this competitive world, people also ask me to do other works, such as, business development work, I openly say, if I need to do it, I will do it in my way, otherwise not.
Further, when you write for you own (costaleconomist.blogspot.com) or when answer some queries online (let it be Linkedin, Quora or Reddit), I feel relieved in this fast and competitive life, as you make yourself update and also make others updated.
I suggest youngsters/fresher, believe in your goals and make your path, as in this world, things change rapidly. I will always give my titles as best example here, in my first corporate company they called me statistical consultant, later, data mining guy, later predictive analytics guy, later machine learning guy, now more or less automation specialist; in all this how I make my path was clear in the rapidly changed world. If you like what you do, developing data science solution, let me SAS/R/Python, you yourself will grab quickly, no need for any push or feel of stress towards transformation.
Q3. How Deep Learning is important to solve the real world problem instead of Machine Learning?
Both exist or develop for solving the real world problems, however, their application depends on the data availability, scalability and other issues you are trying to address.
Let’s take example of email spam detection, it evolved over a period, earlier machine learning now deep learning. However, with respect to loan approval process, still machine learning, they don’t see need of immediate deep learning here.
Q4. What kind of problems have you faced in your Data Science Domain and how did you remove it?
Journey was difficult as most of your colleagues/managers doesn’t understand its preceding and succeeding requirements. They look at me as Magician who has some Magic stick and moreover, they don’t know for which kind of data, data science implementation brings solution to the table.
For example, once a Udupi Hotel business guy reached me for data science solution from demand side, my first question was do you have enough data, after some time they dumped something on me. After looking at that my answer was “your customer decides what he wants from you, not your cook and raw material at your backend, if you have your raw material data then you can only address your supply side issues not demand side issues”.
Hence, now-a-days, I always say ML/DS come into picture only after some data maturity levels, otherwise, don’t even think of ML/DS for now. For instance, you don’t even store, most of your customer data base, then you cannot even think of customer segmentation for marketing.
Q5. You had academic, entrepreneurship and professional experience in your long span of career. Which one do you like most and why?
Entrepreneurship, which always comes with some cost (both from family and financial perspectives)
Q6. You are an Automation Specialist and have a great experience too. How do you see the world after 10 years with full of automation? Will it kill the jobs of humans?
You need humans to make machine learning, machines cannot learn on their own. Automation is more of not about killing jobs of humans, it is more of using latest technologies and moving towards better quality, control and business implementation. There is a famous book, “Small is Beautiful”, written by “E. F. Schumacher”, my takeaway from that is a business of controllable, not compromising on quality with required number of humans is what automation should concentrate on. Move error prone manual works to better automated systems for addressing the needs of human being. I see it as more of structural change as part of technological changes, where people need to move/upgrade towards different technologies.
Q7. Which Machine Learning Model have you used maximum in your career?
Time Series Forecasting Models range from earlier ETS/ARIMA to latest Deep Learning LSTM models. Instead of saying using maximum ML algorithm/model, I have used my brain to understand data more than applying blindly all the ML algos. I am one of firm believers of the statement “all algos/models don’t solve all your business requirements, but, suit for your business requirements, you need to know, which one is that best suits”.
Get more about Dr. Pradeep K Mavuluri @
Ph.D. Thesis: https://shodhganga.inflibnet.ac.in/handle/10603/4190
Capital Account Liberalization in India: Implications for Economic Growth, in 9th UTI Institute of Capital Market Conference: http://www.ssrn.com/link/IICM-9th-2005.html
Revisiting Volume-Volatility Relationship: Evidence from India, Presented at National Conference held at SSS Deemed University, August 2006: http://ssrn.com/abstract=958219
“Measurement of Efficiency of Banks in India”, available at: http://mpra.ub.uni-muenchen.de/17350/