Your DS Story S01E06: Anurag Sharan(British Council)
‘Your DS Story’ is my attempt to bridge the gap between data science professionals & data science aspirants. Here new crop of data scientists will share their experiences, struggles, achievements & their advice so that data science aspirants/enthusiasts can learn and get inspired.
Data science is a superpower for Anurag Sharan. He feels that one can literally see the future of a product and predict events. Every insight he delivers and every business problem he solves gives him a rush, makes him feel powerful. The sense of achievement is overwhelming when his forecast match with the actual with more than 90 percent accuracy.
1. Please tell us a bit about your background?
Hey everyone, I will be sharing my pragmatic view on my journey in becoming a data scientist. Before I begin, I would like to thank Ankit Rathi for taking this nice initiative.
I am an electronic and communication engineering graduate from Amity University. Just like any typical engineer, I got job through on campus placements and I started my job as a database developer at Cognizant Technology Solutions (CTS).
I have worked with TCS and United Health Group (UHG) and currently I am working with British Council as a data scientist. I have been a part of a fraud analytics team to help claim investigators in finding fraud in insurance data by using statistical tests and unsupervised learning techniques. At TCS, I wrote code in SAS to automate CCAR (The Comprehensive Capital Analysis and Review) pre-modelling steps for a bank in US.
Data science is a superpower for me. One can literally see the future of a product and predict events. Every insight I deliver and every business problem is solve gives me a rush, makes me feel powerful. The sense of achievement is overwhelming when my forecast match with the actual with more than 90 percent accuracy.
2. What projects you are working these days?
My main project at British Council is developing and maintaining a time series forecasting model to predict the demand for IELTS in India and China. The recent introduction of new payment modes for IELTS in India and change in immigration policies of Canada and some European countries have made the pattern in which students booked IELTS go haywire.
A mix of time series model, regression models and a few in house mathematical models are used for forecasting. The combination of multiple models is robust solution and adapts to dynamic change is consumer behavior. The forecasting model for India is over 90% accurate in predicting monthly demand which is a huge help for the operations team in maintaining supply.
The analytics team at British Council is trying to scale our model to predict global demand for IELTS. It is a very complicated task because British Councils’ presence is over 109 countries and macro-economic factors will play a big role in governing the demand. We need to establish a right mix of marketing and data science skills to achieve this.
3. How your day to day job looks like?
I will share something which is usually not shared in any data science teaching institute. It is highly unlikely that you will be given a clean analytical data-set and you just have to play with various models. Analytics is fairly new to most companies in India. The data you will encounter might not be suitable for getting any kind statistical inference. This is why command over programming languages is very critical. You might have to write hundreds of lines of code to wrangle your data and making it fit just for an Annova test.
If I had to break my day into parts:
10%: Maintaining the project pipeline.
20%: Helping the stakeholders in consuming analytical insights.
20%: Connecting with business heads to identify pain areas.
50%: Data manipulation and developing and maintaining data models.
4. How you started with DS or transitioned into DS?
I never had fixed career plan for me but I always wanted to be someone who solves problems. The work that I do should have a direct monetary impact and working as a database developer didn’t cut it for me, I wanted more recognition and more sense of achievement. This desire led me to look for such job roles and at that time MBA seemed to be the most suitable option but luck had something else for me in plan.
Like a silver lining, a project at Cognizant required me to learn SAS. It was a great starting point for me to enter the domain of analytics. You could say that I didn’t choose analytics, it chose me. Apart from on job learning, MOOCs on NPTEL helped a lot. I had made a habit of enrolling into at least three MOOCs in every six months and it helped me tremendously in cracking interviews and keeping in pace with the industry.
I very strongly believe that my gradual transition from the role of a data analyst to the role of a data scientist has enabled me in understanding all the parts of an analytics project life cycle.
5. What advice would you like to give to DS starters or DS transitioners?
My advice for data science starters, data science might be just a tool to play with the data but the role of a data scientist is much bigger than it seems. You can make 95% accurate model but it will be worthless for a business if the ROI is insignificant. The whole purpose of an analytics department is to help in making better business decisions. You need a strong understanding of the business drivers before you even think about making a model. For such a thorough understanding of business you need to develop a personal connect with people working on ground. Your mathematical skills won’t get you far if your models do not resonate with the core business goals of your company.
One needs to learn to emphasize, communicate and have a knack of solving problems if he or she wishes to be successful in this domain. In the end, you don’t just play with numbers but you solve real problems of real people. Best of luck everyone.
Thank you for reading my post. I regularly write about Data & Technology on LinkedIn & Medium. If you would like to read my future posts then simply ‘Connect’ or ‘Follow’. Also feel free to listen to me on SoundCloud.