Our Chief Learning Officer, Dr. Anish Roychowdhury, Ph.D. engaged in a conversation with Biswajit Pal , Director of Data Engineering, Analytics & Insights, at TATA CLIQ , who shared some inspiring and intricate experiences, from his journey across Industries and places. Sharing some major takeaways from this chat.

Tell us more about your current role Biswajit. What does a day in your current role look like?

I am heading the Data Engineering Analytics and insight practice for Tata Click. My core focus is on four verticals:

1. The first is around the data operations, which is about the day-to-day management of the data part.
2. The second is about developing a data governance layer on top, basically below the data ops. So, whenever the data is flowing from the source system into the data lake, there is a proper governance layer that tests the data and then the data goes into the lake.
3. The third aspect is about security because it's a commerce platform.
4. The fourth aspect is about the cloud cost optimization. As per our daily activities, we focus on the involvement of Generative AI in our routine tasks nowadays, especially whenever we are carrying out sentiment analysis, since no input is in standard format, we perform these three steps: Language Translation, Summarisation and sentiment Analysis.

You spoke about sentiment analysis and data engineering, which is now an important matter for every business growth, what are some of the major challenges you face regularly?

With innovations and generative AI, though we feel that many conceptualities have been made easy, however, there are many challenges that we face, for example, consider doing sentiment analysis of customer reviews in the context of an Indian market. The reviews are usually in mixed English format and are interrupted with improper diction, and nowadays we have emojis everywhere. So, before we perform the actual operations we must standardise and rationalise it.

I want to emphasize that for any Data Science activities to be fruitful, the core is the data pool. so we deliberately pay attention to enhancing data completeness, data quality and other data aspects. We constantly try to incorporate Data engineering and analytics inside that space. As Data engineers, we also look at historical numbers and sales figures because that helps us build models based on that for forecasting and making recommendations.

How important is domain specialisation? How does your journey across industries have enriched your data science experience?

• To learn about the domain seems very casual initially but being in firms like Walmart has taught me how simple Retail recommendations made from domain knowledge when mixed and backed up with numbers from the Data analytics results are clubbed you will be surprised. For example, the placement of commodities in the supermarket has a huge impact on overall sales, on the one hand, these recommendations are made from knowledge and understanding of the market, and we can analyse and visualise the results for a better outlook.

I came across how scalability is a major aspect because while testing and development we always work on a small sample and then scale the model and trust me, it matters that if we want to implement a recommendation or pricing strategy across 50 or 500 stores across a region or country. We must introduce region-wise amends and restrictions, if any. To summarise be ready to customize your solution as per your stakeholders and end users’ demand.

• Apart from this I would like to share about getting proper exposure in terms of being able to access the major stakeholders. So, being in TCS helped me understand the retail market better and implement the pricing and promotions better. As a retail client, especially in the pricing, and promotion area, ranging from promotional effectiveness i.e., how good this promotion is, then coming to the price optimization side, what should be your optimal price? And, believe me, it sounds very interesting, but it's a very difficult problem to solve given the fact that all these are very foundational to the business. So, if your pricing strategy, goes wrong, your promotion strategy goes wrong. Everything is going to impact your business, the scrutiny that happens.

While hiring the candidates in your team what are the top competencies you are looking for?

1. We are looking for self-driven individuals who have an incremental approach towards problem-solving. See, every problem can be broken into many segments which can be tackled individually and sequentially. We expect and thus test this as a capability in our candidates.
2. Second is good storytelling ability, which means that you should be able to convey your usable solution to both technical as well as non-technical stakeholders alike. This is a simple yet crucial quality to possess as a Data Engineer.
3. Third is the understanding of the data with its solution in the real-time scenario. The piece of code which is working in a Python IDE must be scaled and serve the business value in real-time. Thus, scalability plays a major role in the conversion of the project implementations. As Data engineers, we also expect candidates to pre-process and handle the skewness of the data, if any.

You know about SETU’s PGD course in Data Science & AI. What would be your feedback and suggestions to make it more impactful to the learners?

First things first, it’s a great initiative that the program is preparing well-equipped Data professionals who will understand the real-life scenarios and results. This will help them understand that not every time we rely on and expect ideal parameters e.g., R values, and thus this will fill in the large gap between expectations and actual competency from a trained Data professional. As a suggestion, I feel that scaling of Data solutions can be stressed in this course so that this concept comes in handy as the candidates face business-level problems.

****************************************

Are you aiming for a Career Transition or Career initiation in the Data & and AI Domain, Don't miss out on checking out SETU's PGD course in Data & and AI - one of the most industry-relevant curricula, which offers training better than Job Guarantee. Book a genuine career counselling today.