smartKYC uses AI to help financial institutions speed up and make client onboarding and KYC checks more accurate and cost-efficient. However, it is important that compliance teams are mindful of how the technology they use works as they will need to be able to explain this to regulators when inspections come round. Firms should avoid transferring the responsibility of risk to third parties and must understand the entirety of their due diligence processes. To be on top of an effective due diligence regime, smartKYC recommends that compliance officers consider their:
- Policy
- Process
- Content
- Technology
In this article, we’ll address the technology aspect of due diligence regimes and why it’s important to understand how it works. smartKYC provides a technology-enabled layer to compliance processes that makes due diligence watertight. smartKYC uses AI to analyse structured and unstructured data to corroborate information to provide identity matches scored for relevancy.
- Structured data consists of clearly defined data patterns such as dates of birth.
- Unstructured data is ‘everything else’ and usually consists of information that is not easily searchable such as news articles.
Below is a real-life example of how smartKYC uses AI to identity match analysing both structured and unstructured data:
- There are two well known people called James Caan. One is a famous American actor (D.O.B. 26 March 1940) and the other is a famous Pakistan born British entrepreneur and television personality (D.O.B. 28 December 1960).
- If you were to search just the name James Caan on smartKYC or indeed ask a human to search that name, with no other identifiers, then the results will be an amalgamation on both identities and other people called: James Caan.
- smartKYC finds unstructured data about ‘James Caan’ by way of media archives and gives the data snippets the same likelihood of identity score (66/100), as the only information smartKYC has been given is the first and last name, so as far as the AI is concerned, it could be the same person.
- Now, when we search for James Caan the actor, giving smartKYC the identifiers of his date of birth and citizenship; the same two media snippets are found, but this time the snippet about the actor is scored at 100/100 as smartKYC was able confirm this as the correct identity based on the identifiers it was given for the search and what it finds within the whole document. smartKYC scored this snippet about James Caan the actor as 100/100.
- Within the article, smartKYC finds a perfect name match but also learns about James Caan’s birthday from the same source. This corroborates with his birth date being 26 March, but the article’s year of birth is out by a year, but smartKYC scores it high based on it only being a year out and therefore it is inferred to be the correct identity.
In this example, smartKYC’s technology is not only making matches based on structured data, but using unstructured data to corroborate information and understand more about the identity.
The technology can do in minutes what may take a KYC analyst hours, if not days to search through unstructured data. Due to workload and time pressures, a KYC analyst may match the identity purely on their name and nationality, leading to an incorrect identity match.
The smart technology doesn’t stop there. Using Natural Multi Language Processing, smartKYC’s technology will also review the unstructured data that it finds and tag the record for keywords and will categorise the information by background, related individuals, risk exposure, source of wealth and more.
smartKYC is the world’s most advanced enterprise solution for KYC due diligence automation. smartKYC’s technology drives faster, better and more cost-effective KYC at every stage of the relationship – liberating human effort to focus on decision-making rather than laborious research. Contact us today for more information.