IBM has announced that Signzy Technologies Pvt. Ltd, a financial technology company based in Bangalore, India, is using IBM Watson and IBM Blockchain services to develop a digital trust platform that cuts the identity verification process for financial institutions by as much as 80 percent, improving its ability to comply with regulations and help reduce fraud.

In India, the Know Your Customer norms mandated by the Reserve Bank of India require banks to verify each customer’s identity and provide approved documentation. Banks are under pressure to comply with these requirements, but there is a lack of digital tools to accomplish this efficiently. Most banks are forced to use manual processes that are costly and time consuming.

Signzy is using IBM Watson cognitive computing technology to create a platform that gives banks a comprehensive view of risk. This technology automates assessments that could previously be performed only by humans, potentially cutting verification time from two weeks to two days. The platform complies with national and international requirements for banks to check for fraud, money laundering, financing of terrorism and other illegal activities.

The digital verification process eliminates much of the manual work that was previously required for routine transactions such as opening an account and approving a loan. As a result, banks can reduce the handling and storage of printed and scanned documents, which contain sensitive personal information. Documents can now be stored electronically in the secure Signzy database, helping reduce risk of identity theft and fraud.

Ankit Ratan, co-founder and chief executive officer of Signzy Technologies, stated: “Cognitive computing has the potential to dramatically simplify digital trust processes throughout the financial services industry,”

To confirm identities and look for anomalous behaviour, the system collects both structured and unstructured data, such as customers’ financial history, criminal records, court cases, aliases and government filings, in various formats, and filters out irrelevant information such as divorce proceedings. It then analyses the data in real time to identify connections among individuals and companies.

Banks can upload documents or images into the platform, which can interpret text and understand its meaning using the IBM Watson Document Conversation Service; compare photos to verify that a given person’s documentation matches the individual’s actual appearance using the AlchemyVision Service; and ingest and analyse video content by applying the Speech to Text Service.

To train the system to identify basic patterns that suggest fraud or other illicit activity, Signzy supplied it with common examples of both normal behaviour and red flags. Applying this basic framework to the data, the solution highlights anomalies and generates a forgery score. With its machine learning algorithms, the system continues to learn as it encounters new patterns, expanding its understanding of how fraudulent or criminal behaviour manifests in financial services. Because the Signzy solution uses cognitive computing to facilitate the verification process, it mitigates the risk of human error or corruption.

The platform also establishes a common protocol for running background checks, using the IBM Blockchain service to decentralize the records while keeping the data highly secure. This creates a comprehensive but private digital trail of each customer’s identity, background information and online transactions, minimizing the possibility of tampering.