Tell me about your role at BharatPe and what impact you bring to the organization           

I head the Data Sciences team at BharatPe. I spearhead Data Analytics and Data Science for the company. My role entails ensuring and enhancing data driven decision-making in the organization through insights and reports. Also, I am responsible for building automated and intelligent solutions for decision making using AI/ML. My team works to enable personalized experience for merchants across multiple touch points. The Data Sciences team plays a very critical role for the organization. It helps smartly leverage big data to enable right decision making. Also, it helps drive better customer engagement by powering personalized experience for merchants.

Can you elaborate the role played by AI in powering your products, specifically the marquee ones?

AI plays a very important role in day-to-day business in our organization, across multiple verticals. We use AI for

• Credit decisioning
• Automation of operations processing e.g. automated image and document processing for KYC
• Intelligent systems monitoring
• Improving sales efficiency
• Fraud detection
• Providing personalized experiences to the merchants via marketing (targeted cross-sell and up-sell) and personalized support

As VP of Data Science, what do you reckon are the most commonly faced challenges with financial data in India, especially with SMEs and smaller merchants? How is BharatPe addressing these challenges?

A large chunk of the small merchant segment is yet to go digital. Also, they are not that tech savvy or have a limited understanding of technology. This, in turn, acts as a hindrance to an in-depth data analysis of the digital footprint of the country. Hence, data availability and quality of data are the major challenges with financial data in India.

At BharatPe, we are providing merchants with a quick and easy way to jump on the digital payments bandwagon. Merchants can use BharatPe’s UPI QR and accept payments from a range of payment apps, without any transaction fee. We utilize our payment data and other interaction data to understand the merchant's persona and his/her business potential. The data helps us lend to the merchants, based on their payment transaction flow.

Drawing on your past experience, what are the biggest problems in finance that India should be solving using data? Do you believe fintechs are solving these challenges with data as effectively as possible?

There are a range of problems from finance standpoint for the merchants in India. First and foremost, easy and quick access to financial products and services e.g. payments, loans, micro-insurance/bite sized insurance is one of the key challenges. The access to credit for the underserved and the unbanked is an issue that has been acting as a major roadblock for SMEs for the past many years. Also, an account for the unorganized segment of merchants is another challenge.

I believe that Fintechs are coming up with innovative solutions to serve these needs/ address these problems by leveraging data. They are able to get a more in-depth understanding of merchants and offer them personalized products and solutions.

There is a huge burden on fintechs to ensure financial inclusion using technology, especially for the unserved and underserved. How effective are technologies like AI in truly achieving this? Where are the actual gaps?

I won't say that it's a burden on the fintechs. Instead, I view this as a huge opportunity for the new age fintech companies in the country. There is a large segment of users both on the retail side and B2B side which is currently underserved in terms of financial products. The traditional banking and financial institutions have not been able to reach them/ address their problem.

One of the major challenges is the unavailability of credit history or enough information to take a credit decision for loans. Also, lack of physical presence of traditional banks also acts as a roadblock to driving financial inclusion.

Fintechs are able to bridge the gap as they leverage technologies based on AI/ML. We, at BharatPe, are able to understand credit-worthiness of a merchant, even in absence of a credit history. We utilize data to understand the merchant’s business potential and take credit decisions. AI/ ML driven algorithms have enabled us to give out to thin file merchants who have had no prior credit history.

With a revolutionary product like an interoperable QR code, BharatPe has become a unique player. What other products/services do you envision can cater to the scale of the Indian market?

BharatPe is committed to build a portfolio of financial products that can help small merchants grow their business. In the times to come, we will continue to roll out products in line with the requirements of the merchants. We are working on launching a wholesaler financing product called ‘BharatPe D2R’, which stands for distributor to retailer. We will underwrite wholesalers and extend them a credit line, which distributors can offer to their end-retail partners and kirana owners. We are also working on a range of products in the secured lending space that we will launch in the coming months.

Compared to the demand, the number of data science experts and leaders in India are far lesser. Being one of the few in this pool, what are the most exciting aspects of the job?

The most exciting part of the job is building solutions which can achieve an unprecedented scale and efficiency, e.g. an intelligent chat bot which can understand the context around user queries and automatically provide the response as well as handle 80-90% of the support load without any manual intervention. Only AI can enable such a change in the traditional business process. 

Another similar example is: Improvement in the on-boarding process by automating KYC checks. OCR and Image processing techniques enable an easy on-boarding within minutes and help enhance the customer experience.

What advice do you have for those looking to build a career in data science? 

Focus on fundamentals of machine learning and statistics instead of the next big neural net algorithm released by big techs. In the initial phase of your data science career, you will be applying these fundamentals for 70-80% of your work. Knowledge of SQL and Python/R will go a long way in your data science career. It is important to have a good understanding of the business objective of any problem and how the data science solution will help achieve the same.