Man Vs Machine: How AI-Enabled Verification Can Fetch Faster And Better Results

In a world that is rapidly embracing digitisation and SaaS to decrease both turnaround time and costs while increasing efficiency, the legal and regulatory vertical still remains unautomated.

While we’ve known contracts and legal bonds to be the backbone of any partnership, corporation, or even consumer dealing, it still remains untouched by technology.

However, we are now witnessing a change. Automation and digitisation are slowly making their way into this complex aspect of business, in an effort to make it smoother and easier for the stakeholders involved. In particular, we believe there are two key aspects that should be addressed when it comes to providing a seamless onboarding experience in legal accordance. These include:

• Complete verification and authentication of the documents provided by a customer, and
• Getting a digitally verified signature on the completed contract with a timestamp, which is required to prove the integrity of a data series.

Realising this, we at Signzy created a digital solution focused on delivering an enhanced UX and scalable backend operations, enabling our clients to go for faster digital transformations, while ensuring complete data security and privacy throughout the entire process.

Our clients – mostly banks, NBFCs (Non Banking Financial Institutions), alternative lending institutions, and financial services providers, all require easy and quick onboarding for their customers and we help empower their regulatory processes using Artificial Intelligence (AI), Machine Learning (ML), and Cryptography.

AI-Enabled Verification

To begin with, let’s take the example of generating credit scores for the process of loan origination. A task that required lots of man-hours into calculating credit scores on the basis of a borrower’s financial history, has now turned into an automated task, thanks to deep learning and AI. Hence, what took days earlier, now takes seconds, making both loan approvals and processing a much faster process.

The same applies to verification and authentication of documents when it comes to Know Your Customer (KYC), a factor intrinsic to the onboarding process of any banking and/or non-banking financial institution.

Earlier, a typical process took somewhere between 3-4 days, where an applicant would submit their documents to the institution and the latter would send a person physically to verify the details within the next 3-4 days. This lag did not only add to the processing time but also to the cost of the bank and needless to say, was an inconvenience to the modern day consumer who is used to an ultra-fast process.

To shorten this timeline, we implemented machine learning (ML) and AI algorithms to humanise computer programmes by feeding them with what’s right and what’s not, over a long period of time.

Further, we used deep learning that enables AI to clearly see the distinction between an authentic document and a forged one, even the ones forged on photoshop, something that the human eye might fail to identify.

However, this is not how Signzy started initially. When Signzy was first conceptualised and started, we were looking at making legal contracting a digital matter. We wanted to remove the geographical barriers that caused a hindrance to developing and signing of contracts, and were hoping to develop digital solutions to legal matters as well. Although our solution was readily tried and beta tested by a number of clients and users, they all returned asking for an add-on component which they wanted Signzy to address: end-to-end authentication and verification of their KYC process. Because of this consistent demand, we understood that signing the contract digitally was only the last leg to what appeared to be a more in-depth and detailed process.

Not only was the market open to the idea of digitising legal processes, but it was also asking for more. It was then that we decided to turn Signzy into a full-service solution including video authentication, verification, and contract signing. Based on deep learning and AI, with very little room for error sans bias we created a solution that has now become the armour it was designed to be, to protect banking and non-banking financial institutions from falling into traps of money laundering and at the same time combat terrorism financing.

In recent years, meeting KYC/AML compliances have become a big issue for banking companies. This is primarily owing to the huge fines imposed on banks due to a KYC non-compliance. For instance, on September 4, 2018, Dutch Bank ING was fined $900 Mn for failing to meet Dutch AML compliance. The investigation, conducted by the Netherlands Public Prosecution Service, discovered that the bank failed to execute policies meant to prevent financial-economic crime.

It was this very idea that led us to win the “Payment System Innovation Award” by IDRBT, an RBI initiative. Today, we work with eight major Indian banks and 55 NBFCs including —  SBI, ICICI, HDFC Bank, Mastercard, First Data, Aditya Birla Financial Services, Sodexo, and Edelweiss, among others.

Signzy verifies as many as 2 Mn documents a day, which would traditionally take 3-4 days. At one major bank Signzy’s accuracy rates are higher at 99.1% as compared to the bank’s own manual operations (98.6%).

Finding The Right Cloud Partner To Support The Robust AI Programmes

Keeping AI and ML enabled heavy programmes up and running continuously, especially while supporting large-scale banking clients, we needed a cloud partner who is not only dependable but also one that is flexible with its server availability and payment plans. When we were starting out we evaluated three cloud service providers among which we finally chose DigitalOcean.

Our AI team uses the Droplet (nickname for “cloud server”) product on DigitalOcean, extensively to deploy our core products; KYC and ARI. Spinning up servers on DigitalOcean is really quick. Since we use AI and Machine Learning as base technology for our core products, we need to keep working on multiple computational problems, simultaneously. This requires a lot of server activity and flexible configuration to help scale up processes and start and stop instances, really fast, as and when required. DigitalOcean allows scalability, speed, and flexibility and all this at competitive prices which is a much needed feature for any startup with limited resources upfront. Owing to this, we have been able to price Signzy products competitively.

DigitalOcean’s customer support is also extremely helpful. For a company like Signzy which is modelled to build digital trust protocol, even a moment’s lapse can spell a lot of problems. Continuous support from our cloud partners helps us confidently serve our customers, without worrying about downtime or scalability issues.

As a company whose mission is to establish a digital trust protocol, globally, we need all the computational prowess we can gather, at competitive prices to help increase efficiency and decrease operational costs, and thereby increase our global footprint, rapidly. In our effort to become a company that provides seamless digital identification of individuals globally, we will need to develop, deploy, and work further on our core products. Our cloud partnership with DigitalOcean enables us to envision and roll out our plans as per our scheduled timeline, without worrying about looming costs, downtimes, or scalability issues.

The 3Ms Of The Future

Man, machine, and money have been intertwined for decades; At Signzy, we strive to bring these three together to create a more seamless system of authentication and verification, thereby helping the three function at their optimum capacity.

Consumers are becoming more technology forward, they want faster and more secure processes to meet and exceed the standards they are currently used to. Financial institutions are employing the fastest and most efficient technologies to meet these standards and are also trying their best to cut costs and grow exponentially by beating competition that seems to be mushrooming at a stunning rate.

The only way to add speed and accuracy to these requirements without compromising on quality and security is by adding the much-needed layer of AI and Deep Learning to the entire system. Our team’s understanding of legal principles and contracting, helps us create solutions and products that assist in serving our clients in a better and more professional manner. The possibilities are endless, how the path unfolds is based on how ready man is to accept change.

Note: We at Inc42 take our ethics very seriously. More information about it can be found here.