INTERVIEW: ARNDT GOSSMANN, DGTAL
DGTAL unveils new AI tool, Grabber, ahead of Baden-Baden
Grabber is an intelligent processing tool that extracts information from semi-structured documents.
DGTAL, a Germany-based software company which uses artificial intelligence (AI) to give insurers better insight into their claims data, has announced its latest product ahead of the Baden-Baden Reinsurance Meeting.
The tool, called Grabber, complements its earlier product Driller, which helps companies make sense of masses of unstructured data.
“If Driller is like a satellite that gives you a great overview of a mass of claims files, Grabber is like a drone that is able to pick specific information in claims documents,” says DGTAL chief executive officer Arndt Gossmann. “It grabs the relevant information.”
Grabber is an “intelligent processing tool” that extracts information from semi-structured documents such as bank statements, invoices, quotes or other forms.
For example, it can pick out specific data from a set of health insurance invoices and allow the user to compare the data. “It knows where to find it and where to put it,” says Gossmann.
The launch of Grabber comes a few months after DGTAL announced a €3 million ($3.3 million) investment by the venture capital arm of German insurer VGH Versicherungen, several family offices and Gossmann & Cie, a firm co-owned by Gossmann which provides strategic solutions for managing insurance liabilities and regulatory capital.
Driller and Grabber both help companies get better insights from the mass of unstructured data that is in their possession but not stored in structured databases and therefore inaccessible to conventional means of data analysis.
WATCH THE FULL INTERVIEW HERE
“AI significantly helps insurance companies improve daily operations.”
Arndt Gossmann, DGTAL
Gossmann said that Deutsche Bundesbank, Germany’s central bank, estimates that unstructured information accounts for 80 percent of data gathered by companies, compared with only 20 percent that is structured. Management consultant firm McKinsey & Co has estimated that the split in insurance claims files data is even more pronounced: 98 percent unstructured data and only 2 percent structured.
As a result, Gossmann believes, there is plenty of room for AI companies to grow. “I’m not worried about market potential for whatever type of solution that helps to grab or to drill unstructured information. DGTAL has gained two global, blue-chip clients in the insurance sector for Driller since the recent funding.
“ΑΙ significantly helps insurance companies improve daily operations and reduce leakage and fraud,” he says. “It’s not only a nice techie software—it delivers results for those that apply.”
The machine is learning
According to Gossmann, the value of Driller increases the more it is used, as it is self-learning.
“With every claims file that goes into the machine, the machine is getting better at reading and assessing it.” The result, he says is “more answers out of the same data”.
Gossmann says the company’s current focus is on solidifying the platform so it can run multiple portfolio assessments at the same time. More broadly, DGTAL is targeting breakeven by the end of 2024.
Gossmann is not concerned by European regulatory action governing AI, which has the potential to slow down development of the new technology.
“It should allow for much more thorough pricing negotiation.”
“The attitude of regulators is extremely open to AI,” he says. “They are on top of things. From a regulatory perspective, there’s a clear expectation that AI will be used to improve businesses.”
That is a good thing for customers, he adds. “If you have better information you’ll be better in management, policyholders are better protected, and risk assessments are better.”
Breaking down and using the mass of unstructured data is key to that process. “If you have four times more information available, it’s almost impossible not to improve.”
Ahead of the Baden-Baden meeting, Gossmann says reinsurance will be a direct beneficiary. “If you are able to deliver better and more in-depth information as a direct insurance company to your reinsurer, this creates a huge benefit because it should allow for much more thorough pricing negotiation.”
For example, he says, you can work out the specific effect of inflation in different portfolios and its impact on pricing, rather than relying on an overall inflation rate plus risk buffer. “That’s good for the reinsurer, and good for the direct insurer.
“The bottom line is that the insurance business is based on information. AI is made to deal with information,” says Gossmann. The first wave of AI has been taking advantage of better information, making better and faster conclusions, he says.
“In the next stage, we will learn how to use the technology to change our operational models as insurance companies. That’s a longer, but very exciting process.”
Arndt Gossmann is the chief executive officer of DGTAL. He can be contacted at: a.gossmann@dgtal.io
Main image: Shutterstock / Rawpixel.com
