Building an analytics team from scratch at Bharti Axa Life Insurance

Neharika Agarwal Bharti Axa

Neharika Agarwal has 12 years of experience in connecting data, analytics and technology to drive decision making, strategy development and implementation in the areas of revenue growth, operational efficiency and new product development across multiple industries such as Financial Services and Consumer Products. She started her career in Quant Equity Research at Morgan Stanley Capital International as an analyst. She is currently working as a Vice President and the Head of Analytics and Business Intelligence at Bharti Axa Life Insurance, where she has built a data analytics team from the ground up.

Analytics India Magazine caught up with her for an interview.

Edited excerpts:

AIM: How did you start your journey in the field of data analytics?

Neharika Agarwal: Interestingly, I am not a data scientist by education. I come from a financial background. During the time I was pursuing management studies(with a focus on finance) at BITS Pilani, I got an opportunity to intern with the Quantitative Equity Research team at Morgan Stanley Capital International for six months. It was my first formal brush with analytics and data science when it wasn’t even a recognized industry word. Over the years, I complimented my finance degree further with a CFA and FRM charter holder. I reckon that it’s actually my quantitative finance background which has helped me develop a keen understanding of data, application of data and technology to solve business problems, an analytical bent of mind and a curious and solution-oriented mindset.

It’s important to understand that data analytics is nothing but analyzing raw data, organizing it, finding hidden patterns, and unseen trends, interpreting it to derive valuable business insights and communicating them to businesses. The size and the type of data analyzed vary from a couple of kbs to massive big datasets (running into terabytes), from structured sales-type data to unstructured customer interaction data. Data manipulation and insight generation can be done using basic MS-excel using classical statistics or machine learning technologies such as neural networks, NLP, sentiment analysis etc. It is important to understand that technology is just an enabler to data science/data analytics professionals which helps in drawing meaningful conclusions from complex and varied data sources, now possible by advances in parallel processing and cheap computational power. Without understanding the application and conversion into business-speak, technology in itself is meaningless.

In several companies, the data science team doesn’t even sit within the technology vertical anymore. In my current role, I report to the CFO directly which enables us to impact all business functions.

AIM: Tell us about your team, a relatively newer addition at Bharti AXA.

Neharika Agarwal: This team was set up in April 2020 post our CEO, Parag Raja, took over the reins of Bharti Axa. His vision was to create a Center of Excellence for a data-enabled decision making culture within the organization. I currently have a 13 member team reporting to me with three members in Core Data science and the remaining in reporting/business intelligence. I have deployed a hybrid operating model for data science projects currently where apart from my in-house team, I also partner with multiple specialized organizations on specific use cases. In the last two years, we have successfully developed and deployed multiple ends to end ML enabled solutions at scale to impact revenue, persistency and cost efficiencies. On the Business Intelligence and reporting side, I manage the entire daily sales reporting for the complete hierarchy, performance management and rewards and recognitions, including short term insight generation projects on a day to day basis.


AIM: You built the analytics team at Bharti AXA from scratch. How did you pull that off?

Neharika Agarwal: There are broadly three challenges that are usually encountered while setting up a data and analytics team in any organization. Firstly, setting up an internal analytics team is usually an organizational challenge due to a lack of executive sponsorship. The problem is big data or analytics is often used as a catchphrase – people would expect immediate-short time gains, and failure to deliver would lead to analytics being written off. Businesses are often worried about “what if it impacts my business negatively?” Thankfully, I did not face this challenge as I did have complete executive sponsorship given that the unit was created by the CEO and housed under the CFO. However, creating a balance between the immediate goals of the organization and the long term capability building including gaining the confidence of the business heads required delicate handling. The actual goal was to create a data-based decision-making culture within the organization, which is a long-term journey. To establish business confidence, we focused on demonstrating proof of concept with smaller projects with quicker outcomes/ROI/ positive impact on business goals. And in parallel, we continued to work on core data science projects with a 1-2 year gestation period for green shoots. For example, while there are complex ML models behind, we have now been able to create a common language spoken by all business heads for assessing the quality of customers and what actionables should be taken to improve the overall customer experience and business. The biggest learning in my analytics journey has been to learn to focus on “why” and “how” and convert any output to business language.

The second challenge usually is the quality, veracity and unavailability of data in disparate data sources including inconsistent ways of interpretation of the same data. Bharti Axa was not immune to this problem. To solve this, I partnered with the CTO and his team to create a Data Lake which serves as the single source of the data across the organization. Also, given that I’m part of the finance function, we have also been able to standardize the definitions of all business KPIs and their interpretation.

The third challenge was to hire the right talent pool including managing the candidates’ expectations and the business realities of a fledgling analytics team. I usually focus on evaluating the business acumen, problem-solving skills and the can-do attitude of the candidate among technical skills. To overcome the hiring challenges, as mentioned earlier, I have a hybrid model where my internal team focuses on urgent and important projects while I work with partners on more long term capability building.

AIM: How did you recruit for your team? What advice would you give to recruiters selecting candidates for the analytics role?

Neharika Agarwal: All said and done, I am not going to take away the importance of the technical skills that are required for an analytics role. In fact, it is almost akin to hygiene. Candidates must be well versed in a couple of coding languages ​​– Python, R or SAS. They should also be hands-on with visualization tools, so when we are looking at an analytics project lifecycle, one must be able to visualize the output and connect it to different KPIs. This is mandatory to drive the adoption of whatever project has been implemented.

Second is knowing the math behind a project as that helps in making connections to the kind of business problem that you are trying to solve. Here, a statistics background helps immensely.

Third, I think a lot more focus has to be on real-world applications of the data science techniques. Education institutions need to provide a 360-degree view of business and how whatever you are doing is linked to solving a particular business problem. Hence, installing that mindset is extremely important for a successful data science professional.

Source link


What do you think?

5k Points
Upvote Downvote

Written by admin

Leave a Reply

Your email address will not be published.

GIPHY App Key not set. Please check settings

Weaker earnings foreseen for US life insurers in Q1

In review: the essentials of insurance and reinsurance law in Cambodia