NEW DELHI, Feb 22 (PTI): When an ailing 80-year-old couldn’t go to his bank for the annual verification of documents necessary to get his pension, HyperVerge’s video KYC solution at the State Bank of India stepped in. He was verified remotely in under two minutes and soon received a confirmation SMS.
“Such powerful, heartwarming stories drive us to do better and innovate every day,” Kedar Kulkarni, CEO of the AI-based identity verification solutions company, told PTI in an interview.
The HyperVerge journey started in 2014 when a group of IIT-Madras students with expertise in computer vision and machine learning and a “strong desire to solve real problems” came together, he said.
Related Posts:
The founders started with the dream of becoming “the next Facebook or Instagram” but soon realised the technology had more powerful applications in enterprises. The product vision, Kulkarni explained, is to democratise digital transactions by making them fast, easy and cheap.
Over the last three years, HyperVerge has onboarded 700 million users, Kulkarni said, adding that large enterprises had started using HyperVerge AI engines for verifying identities of their customers by 2018.
“Billions of people today do not have access to basic services such as savings, credit, insurance, and investments. Every adult should have this access. We are building technology products to accelerate this change,” Kulkarni said.
Excerpts from the interview: Q: What is the story of HyperVerge? What have been the key milestones in the organisation’s journey so far? A: In 2014, a group of IIT Madras students with expertise in computer vision and machine learning with a strong desire to solve real problems for people came together to form HyperVerge. We raised a seed round of USD 1 million from NEA, Milliways Ventures and Dallas Venture Capital in 2014.
We started by building consumer applications with the dream of becoming the next Facebook or Instagram. But eventually, after a few iterations, we realised that our technology had more powerful applications and value for enterprises and pivoted to a B2B business model in late 2017.
In 2018, large enterprises started using HyperVerge AI engines to verify the identities of their customers. Now, the team has grown to 160-plus members and we have increased our presence across multiple geographies. We are profitable and our customers have fuelled our growth so far.
HyperVerge’s technology competes with some of the biggest, largest, and most heavily funded companies and stands out as one of the best in the world.
We have onboarded 700 million users over the last three years. Leading enterprises such as Vodafone, Jio, SBI, LIC, CRED, Home Credit, L&T Financial Services, Swiggy, Grab and others use HyperVerge for seamlessly onboarding new customers and business partners, authenticating existing customers and agent, and identifying and stopping fraudsters.
These clients choose HyperVerge because of our ability to handle massive scale, our globally top-ranked AI engine, and our ability to make cutting-edge AI work on low-end phones running on 2G and 3G networks. We are now expanding globally with clients across telecom, digital lending, neobanks, crypto, gaming, sharing economy, logistics, insurance, ed-tech, banks, fintechs among others.
Q: What is the mission and vision of HyperVerge?
A: Our product vision is to democratise digital transactions by making them trusted, fast, easy, and super cheap. Billions of people today do not have access to basic services such as savings, credit, insurance, and investments. Every adult should have this access. We are building technology products to accelerate this change.
While we built a strong economic engine in the form of HyperVerge’s identity business, we also wanted to have a very meaningful contribution engine. As early steps towards this journey, we have set up HyperVerge Academy (HVA).
HVA is focused on uplifting students from low-income families by upskilling them on in-demand technology skills. We have trained for free and placed 60-plus learners in software companies, thereby increasing the average income of these families.
Q: What has been your experience working with large enterprises?
A: I would like to talk about one impactful story here. Recently, an 80-year-old wanted to receive his pension but, due to an ailment, was not able to go to the bank to verify that he is alive. Our video KYC solution in SBI was used to verify him remotely in less than two minutes and within a day, the customer received an SMS saying his life certificate is verified. Such powerful, heartwarming stories drive us to do better and innovate every day.
Our work with our enterprise partners has been very fulfilling and rewarding. Every customer’s challenges and requirements are different even if the companies are direct competitors of each other. We work very closely with our enterprise clients to solve their specific challenges enabling them to achieve higher business impact.
Q: Identity verification is a crowded space with multiple players. How do you differentiate your offerings?
A: Though there are multiple players in this space, HyperVerge stands out in a few aspects.
Global players who compete with us are largely system integrators of outsourced technology components. We build in-house tech and are significantly ahead on all AI accuracy parameters leading to much higher value add in all the geographies we compete in.
Our AI Accuracy on OCR (optical character recognition), face recognition, and liveness is the highest among competitors. We have made breakthroughs in our OCR technology and we can now support OCR for new documents with as low as 10 training samples with very high accuracies.
Our technology can be deployed in low-bandwidth areas (even in 2G). While our competition caters to end users with high-end devices (cost more than USD 250), our tech supports about one million device models (that cost less than USD 150 each).
Our solution can be integrated very quickly and with very minimal developer bandwidth.
Q: What technology infrastructure do you use for your products?
A: HyperVerge’s product offerings include cloud-based APIs and edge solutions that work on smartphones. We have partnered with Amazon Web Services (AWS) for our technology infrastructure. On the cloud, our product is hosted mostly on autoscaling systems, along with other auxiliary services including Amazon Simple Storage Service (S3), Amazon Simple Queue Service (SQS), AWS Web Application Firewall, Amazon Relational Database Service, AWS Lambda for serverless compute, etc .
Since our AI inference APIs run multiple neural networks, Amazon G4DN based GPU-powered Virtual Machines form a core part of our infrastructure.
Q: How has cloud technology and AWS enabled you in your operations?
A: AWS has helped us scale seamlessly to serve much of the fintech ecosystem in the country and across the world. Its ecosystem of products makes it easier for us to work as a small nimble team while building comprehensive and robust solutions. The large ecosystem of talent and online communities has enabled our teams to build very reliable systems for high volumes. The availability of data centres across the globe has been a key enabler for us to serve clients from across the globe.
