Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

It Started with Jungle Animals: Now Onfido is Solving Identity


London-based company Onfido is on a quest to help anyone prove who they are. It sounds easy, but it is in fact a task of gargantuan proportions.

When a person needs to demonstrate their identity—for instance, when opening a bank account— that usually involves a trip to a physical location and a face-to-face verification from a clerk glancing at their ID card.

Problem is, Onfido CEO Husayn Kassai explains, people in many parts of the world cannot rely on offline infrastructures (that is: bank offices) and in the same breath they cannot access online financial services without proving their identities first.

“The current process of identity verification is a big issue for underbanked and unbanked people,” Kassai tells SHACK15 News. “It’s also an insecure process: almost two trillion dollars every year are laundered as a result of identity frauds.”

That is the problem Onfido has set out to solve, by harnessing machine learning and computer vision.

Onfido can verify somebody’s identity using only a couple of photographs. Users are asked to take a pic of their photo ID and then a selfie— allowing Onfido’s software to compare the two images and ascertain they match.

To ensure that the selfie is genuine, the user  has to carry out some random tasks—such as turning their head, or blinking— while taking the picture.

“They don’t know what we’re gonna ask them, so we know there is an actual person there,” Kassai says. “Obviously, right know we can’t be sure if they are doing it willingly: we don’t know if there’s somebody there with a gun asking them to do submit their details.”

The system’s computer vision algorithms can also automatically assess whether the document being shown is genuine or fraudulent—thanks to a database of over 600 document types— and the company periodically shares with the authorities the images of forged IDs submitted on the platform. (In all other cases, both the photo of the document and the selfie are deleted from Onfido’s database.)

The company was founded in 2012 by Oxford students Ruhul Amin, Husayn Kassai, and Eamon Jubbawy. The trio built upon a technology Amin had designed for his university thesis: a computer vision tool able to spot wildlife in the jungle.

“We were thinking about a business idea and when we looked at verifying documents [Amin] said : ‘this is such a classic computer vision problem’.”

Being among the early movers in the field allowed Onfido to access vast amounts of images , and to further improve its algorithm; recent advances in biometrics have also contributed to make its technology even more sharp-eyed.

Onfido’s technology currently powers the verification systems of over 1,500 businesses worldwide— including UberEATS. Until recently, it helped Uber proper with cab drivers’ background checks in London; Uber was  eventually requested to use the services of another firm with which Transport for London had an exclusivity deal.

Just weeks ago, the 150-employee company raised £22 million in a Series C round led by Crane Venture Partners and backed by Microsoft Ventures, and Salesforce Ventures. Onfido has so far raised about £45 million ($60.3m) in seven investment rounds.

Kassai explains that  the fresh cash will go into enhancing Onfido’s technology and expanding its international reach.

“We are looking forward to expanding to new regions,” he says. “Our core markets right now are the US, the UK, and  India; we’re now expanding to the Far East, Latin America.”

Images via Onfido


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter