While the age of the internet has provided us with more accessibility to services, it has also given rise to an increasing pool of fraudsters. That’s precisely why it is even more pertinent today to verify the background and identify suspicious elements at source, so that organisations can engage with genuine entities.
Founded in 2011 by Ashok Hariharan, and Vineet Jawa, Mumbai-based IDfy does exactly that through their latest machine-learning-based anomaly detection, machine vision, and identity authentication techniques. IDfy’s mission is to unlock opportunities for genuine people – to date, their platform has touched the lives of 150 million people.
Ashok, along with Ashish Sahni, who helms the role of the Chief Technology Officer, speak about IDfy’s unique technology, the impact they are driving in this domain, and how their collaboration with AWS has been fruitful for their growth.
The mission
IDfy has been consistently working towards helping companies onboard individuals and enterprises across use cases and industries like employment, financial services, and the gig economy.
“Most people who are digitally savvy have already used us. Our mission is to authenticate as many people and enterprises. We sit behind several onboarding journeys, be it wallets, real money gaming, employment, banking, FMCG – and are helping the largest players in that industry onboard genuine people,“ revealed Ashok.
The platform filters out fraudsters, using their cutting-edge machine learning/artificial intelligence (ML/AI) models within seconds, to unlock opportunities for genuine people.
How IDfy works
The Identity Verification platform provides services like document proofing, document information extraction, assessing face quality, data verification, and other sets of use cases to carry out the authentication, thereby helping to know a customer better.
“This enables our customers to integrate this platform with their systems to do seamless account opening, withdraw money, loan approvals, and a host of related use cases that they are looking to solve,“ shared Ashish, adding that this is where AI steps in to design custom ML models that are optimised for accuracy and turnaround time.
The architecture for IDfy has been built on a set of horizontal and vertical capabilities; most services of which are developed using open source technologies.
“With this approach, we have gone and built higher-level services such as rule-based orchestration engines and video services to enable our customers to build end-to-end customer journeys. Moreover, our custom design ML models utilise Amazon Rekognition hosted on the AWS platform,“ said Ashish, adding that some of their unique approaches like instrumentation-based development methodology provide a feedback loop that is used to improve their product engines.
Scaling up with AWS
IDfy, to its credit, enables 25 million authentications a month, sometimes handling as many as 10,000 authentications per second. This requires them to be able to scale dynamically to handle unusual spikes and deal with quick turnaround on their volumes.
“We use Amazon for scaling up and down our services, as well as for AI/ML capabilities of Amazon Rekognition, which we use for data labelling and sentiment analysis. We have had a true partnership with AWS, both as customers and vendors for over six years,“ added Ashok, sharing that AWS has helped them train their in-house models with high technological capabilities.
The business impact of the contribution driven through Amazon AI/ML services is significantly high. That’s because all of their services use some or the other aspect of AWS.
Looking into the future
IDfy has ambitious business growth plans across industries and aims to expand to other geographical markets.
“We want to add new AI-based machine learning models to our platform. In order to achieve this, we want to work with AWS to help bring AI/ML capabilities that will enable us to build the use cases in areas such as risk and fraud and video analytics use cases,“ he concluded.
