Darwinium deploys on a Content Delivery Network at the edge, and is designed to provide full visibility to both security and fraud detection teams.

Today San Francisco-based Darwinium, which makes an edge-based digital security and fraud prevention platform, is announcing that it has raised a further $18 million in Series A financing, bringing the total investment in the company to $26 million. The money will be used primarily to finance Darwinium’s global expansion.

Darwinium puts security and fraud prevention to the edge, deploying at the CDN [Content Delivery Network] level instead of using an API-based solution. It utilizes streaming algorithms there to provide complete visibility and control of every digital interaction in real time, to both fraud detection and security teams.

“Darwinium takes an evolutionary approach,” said Alisdair Faulkner, co-founder and CEO of Darwinium. “In a world where machines attack humans and machines attack machines, AI will blow apart everything and fraud security has to be completely rethought.”

Faulkner noted that AI is now being used to test websites for business model vulnerabilities.

“This might allow them to determine thresholds, like a company not checking transactions under $1000,” he said. “Previously they would tend to use ‘low and slow’ attacks to stay under the radar, but now they can run combinations of DDoS attacks with others, because of this improved intelligence.”

Company responses tend to be less effective because fraud detection and security teams have typically been in two separate siloes.

“Fraud is more likely to get through because these teams don’t talk well to each other,” Faulkner said. “Without the two sharing a single layer of visibility, they will will chase their tails. We attack this rather than take it piecemeal, using the existing Content Delivery Network infrastructure. This lets us process it close to the customer, reducing latency. And it lets both teams see what is happening and differentiate between real customers who do a lot of transactions, who are typically their better customers, and fraudsters. This means that the good customers aren’t slowed down, while the bad guys are stopped. 90% of good fraud detection is good customer recognition.”

Faulkner said that another advantage of deploying at the CDN is that Darwinium doesn’t provide a single point of failure.

“These tasks have been typically done by API-based products that risk-assess moments-in-time, he stated. “We run on your CDN as it receives traffic, using streaming algorithms. “DevSecOps is enabled to have full control and visibility which can be split with engineering and security. The data science teams can also  continuously change and update models without having to go through full engineering cycles.”

While 2023 has seen vendor after vendor announcing new Generative AI solutions, Faulkner said that Darwinium is already well positioned in this market.

“We were originally focused on digital risk when we were formed in 2020, and we saw this coming even before that,” he said. “What opened my eyes was AlphaGo, software that beat every Go champion, and which was not taught by a single human. This would be the way of the future, and we were created with that vision in mind.”

Faulkner also noted that Darwinium was also able to do well over the past several years before Generative AI took off, when the reputation of AI technology was decidedly more mixed.

“Many CEOs used to say that AI was just ticket generating machines, but we don’t sell them something that stands in isolation that is just another systems integration product,” he said. “The CDN deployment lets us normalize all the data. That’s the best and only place to get this job done. We aren’t selling a big bang.”

Darwinium’s platform is built for eCommerce merchants, online marketplaces, fintech providers, gaming and gambling operators and retail banks.

“Financial services companies are our biggest customers, followed by marketplaces, e-tailers, gaming and government,” Faulkner said. “Banks take longer to install and like to see more proof points. eCommerce lives and dies depending how they service their customers. eCommerce is typically a beachhead use case for us, and then financial services.  We are built for a world where ChatGPT will be automated to do transactions, and banks can let it all through or none. In that environment, you have to have the equivalent of a CCT camera looking in real time at what a person will do, to determine the equivalent of are they going to the safe or to the fridge.”

Darwinium has several angles for channel partners.

“We have an ecosystem around our no-code platform and a marketplace which lets vendors be dropped in,” Faulkner said. “We don’t resell these integrations. We are also an enabler for the upgrade of security detection. The channel might have sold a solution, but getting it through engineering becomes the bottleneck. We can leverage third party engineering. CDN is also a channel for us. Partners who build risk models for customers, who do business consulting and technology, are a great fit, as are dedicated security folks who sell holistic fraud and security solutions.”

The funding round was led by U.S. Venture Partners (USVP), with participation from seed investors Blackbird, Airtree Ventures and Accomplice.