AI crucial in preventing security attacks, fraud: Razorpay co-founder

In a Q&A, Razorpay co-founder talks about opportunities that pandemic has thrown up for them to rejig their business and leverage digital technologies to serve customers

The Covid-19 pandemic has hit small businesses hard, and fintech unicorn Razorpay is helping them to bounce back through digitisation. Shashank Kumar, co-founder and chief technology officer, Razorpay, tells Peerzada Abrar and Deepsekhar Choudhary in an interview that the health crisis has thrown up opportunities for them to rejig their business and leverage digital technologies to serve customers. Edited excerpts:

How has the digital payments ecosystem, including Razorpay, evolved since its nascency?

When we started, the payments infrastructure itself wasn’t in a good place. We had to do transactions on 2G internet back in 2014-15. Then you had a mix of payment options across net banking, cards, and later UPI (Unified Payments Interface). We were able to focus on building the payments infrastructure and later started building technology products on top of the payments infrastructure. We launched products like payment links, invoices, subscription management and B2B payments. We built products which made it very easy for anyone to go online and start accepting payments without doing any technical integration.

We call this space “no code apps”. You just register on our website and are ready to accept payments within minutes. You don’t need to rely on any developer to do so. We are now a lot more focused on creating trust in digital commerce and building the right incentives. How do you identify genuine sellers from fake sellers, and figure out whether the seller will provide you with a good experience? We can help small merchants acquire a lot of new customers through our platform and then push them to transact digitally. All these are enabled by artificial intelligence (AI), machine learning (ML) and very good fraud management.

What technology bets is Razorpay making over the next few years?

One core focus is transaction success rates, and we will keep investing on that front. UPI has played a huge role on that front. Affordability is going to be another major area where technology will play a huge role. There are multiple initiatives that we are driving there. As internet penetration increases further, financial penetration also needs to keep pace. As we enable more sellers in the financial space, how do we build the right set of products for them?

We are betting heavily on neo-banking as a product segment, because we feel businesses should get the same level of personalised experience on banking as you get on Instagram or Facebook. There is also a huge credit gap that exists in the country. How we bring credit to a majority of SMEs in the country is going to be a huge focus area for us.

Last year Razorpay launched tokenisation to help businesses comply with RBI guidelines. How is the industry adopting such solutions?

There’s no option but to adopt, otherwise, you will give a significantly inferior experience to customers. We expect that enterprises will comply with the regulation and go live with tokenisation. RBI has given a six-month extension, keeping in mind that this is really important. There are very few companies that have built a very good tokenisation stack. We have taken a lot of large-scale customers on the tokenisation platform. Some of them are ramping it up across their customer base. We have also launched a tokenisation product for banks, and are helping them understand the compliances.

The increase in online payments has also increased cybersecurity and privacy issues. How is Razorpay addressing these issues?

One is on the consumer side. How can we better educate consumers not to go to websites where there are fake merchants who will just take your money and run? The other is the merchant side. How do you prevent merchants from misusing the platform, creating fraudulent transactions and duping customers? We have invested a lot in these areas at various levels. We monitor it very aggressively.

What is the role of AI and ML in addressing such issues?

Machine learning can play a huge role. We have a good amount of data to identify how frauds happened in the past and how we can prevent similar patterns from getting repeated. We can also identify new patterns and clamp down on those. We have also been partnering with all law enforcement agencies. We are investing a lot from a security perspective. There are a bunch of audits that we do on a monthly basis through various partners. We run a lot of simulations on our system to better see how our systems are able to stand up to external attacks.

Right from onboarding merchants, transaction monitoring, underwriting, to preventing security attacks and giving better support experience, in all these places, AI and ML play a critical role. AI is one of the new frontiers of technology that helps us perform a lot of operations at scale and at lower cost. In all the operations that we do at scale, including better transactions, success rate or preventing fraud, AI and ML play a significant role in helping us build better models.

How do you see the third wave of the Covid-19 pandemic impacting the industry?

The pandemic is hitting small businesses and enterprises harder. As an organisation, we are focused on how we can help small businesses to tide over the impact. How can we help people transform to digital, understand the value of the internet and online transactions, and not do things the old way?

The pandemic is accelerating the digitisation trend, and businesses that are not transforming will be left behind. As an ecosystem, we are slightly better prepared for the pandemic this time. You should look at it as an opportunity -- how you can redo your business in the new age, leverage digital technologies and figure out new ways of serving customers? We have built a programme called Razorpay Rize, where we want to incubate one lakh digital startups.