With AI, this startup wants to help other companies spot fraudsters

Formotiv, based in Center City, raised a $650,000 seed round this summer. It promises to help companies in financial services and insurance spaces identify risky behavior.

Smart911 is now in Kent County.

(Photo by Flickr user Neon Tommy, used under a Creative Commons license)

Center City-based Formotiv, a B2B startup founded in 2016, built an artificial intelligence and machine learning platform they say can help companies in financial services and insurance spaces spot risky behavior online.

Per Formotiv partner Bill Conners, an alum of Radnor-based Relay who’s now a partner in the startup, the tech platform works like this: Using information from customer’s digital behavior while interacting with online forms (time to complete applications, repeated changes to responses and other red flags), the company identifies potentially risky client profiles and recommends additional steps to verify the information.

But there’s a bit more to the pitch.

“The advantage is two-fold,” Conners said. “We’re helping companies improve their digital apps by identifying pain points in their application process, and in real time we can recommend ways to pull their customers through the online process. On the other end, we can help detect possible risky or fraudulent behavior.”

With fuel from a $650,000 seed round closed last August — which saw participation from a mix of angel investors from the tech and VC worlds — the company recently opened an office at Spaces in the Hale Building. Formotiv employs 10 full-timers, with four based in Philadelphia and the rest in an off-shore dev hub in Vietnam.

Is this system unfair? We’ve heard form people using accessibility tools (like screen readers) that certain apps will time them out while they’re still trying to complete the process. Senior citizens also tend to be slower adopters of tech, which might mean they’re also slower to complete online processes.

The partner said the difference here is that the company’s tech won’t make anyone get rejected outright for taking too long, or making mistakes on a form. It will simply recommend extra steps — like answering additional questions or providing IDs — to qualify users responses before going through automated application forms.


“We do think the process is fair,” Conners said. “People can make mistakes. That’s going to happen.”

(One machine learning algorithm is helping call center workers at Philly’s Benefits Data Trust better assist potential SNAP clients.)

The company, founded by software developer Andrew Schwabe, has five paying customers to date, mostly in the financial services and insurance spaces.

Subscribe to our Newsletters
Technically Media
Connect with companies from the community
New call-to-action