In a couple of weeks, we’ll participate in The Big Score in the lovely city of Ghent, Belgium, an international venture capital & corporate innovation event that will feature 50 of Europe’s most promising growth stage scale-ups and 25 interesting Belgian startups.

One of the scale-ups that will be also in attendance is Berlin-based omni:us. The company just received a €1.6 million grant from the European Commission’s European Innovation Council (EIC), and we thought their story was super interesting, so we’re featuring an interview with Sofie Quidenus-Wahlforss, one of the co-founders and CEO of omni:us.

1) When, where and by whom was the company started, but most importantly: why?

Omni:us was founded in Berlin in 2015 by Eric Pfarl, Stephan Dorfmeister, Harald Gölles,
Martin Micko and myself. But the roots run deeper. My previous company developed
automated processes for digitising documents. At some point it became clear to me that a
new problem was emerging, namely a myriad of documents that were not machine-readable.

This was the kickstart behind the idea of omni:us, a powerful AI that can read, process and understand a wide variety of documents. When we had the technology, we looked around for where there was the most pressing need for our solution and quickly came across the
insurance industry. Insurance companies still have to process on a daily basis a multitude of different, inconsistent and often handwritten documents.

2) On your website, you are described as an “AI-first company” – what does that mean concretely?

AI-first means to us that any problem can be solved by AI. If enough data points are
available and the AI has been trained accordingly, it can not only solve problems, but also provide well-founded decision-making aids for the responsible employee. This saves a considerable amount of time, which in turn can be invested into direct customer service to better advise and assist them. There is an especially prudent use case in claims management, where processes are historically rigid, inefficient and often too impersonal.

With an AI-first strategy, large increases in efficiency become possible. This leads to faster, more individual and empathic resolution of insurance claims – a service that is becoming increasingly important for customers.

3) Would you say it’s early days for AI in the insurance industry, and could you
describe how big the potential is, and why?

We are certainly just at the beginning. There is huge interest in the technology across the industry, but it is only now that implementation is really getting underway. There are still a few concerns to be cleared up here and there. But the potential is vast. AI is already unbeatable today when it comes to processing repetitive, computation- and storage- intensive tasks – and insurance companies operate around an incredible number of such tasks.

4) What are some interesting applications of AI that you’ve seen, whether the solution
was provided by omni:us or not, specific to the insurance sector that makes it
faster/better/stronger/more efficient?

There are already some exciting and promising approaches. InsurTech FRISS, for example,
has developed AI-based solutions for risk and fraud detection. This makes it possible to automatically check damage reports for attempted fraud. The degree of suspicion is then clearly displayed in the ‘FRISS score’.

There are also interesting approaches for the assessment of damages. AI-supported image analysis methods can now be used to determine with astonishing accuracy how much repair costs will be incurred, for example after a car accident, without an on-site expert having to inspect the damage.

And, of course, document recognition and data extraction as we do it is also a potential game changer. Technologically, we are cracking really tough nuts.

5) On your website, you boast about your technological prowess and domain expertise. For the latter, which specific experience do the people in your company have with regards to the insurance industry in Germany and beyond?

We have a number of employees at different levels who have many years of experience in
the industry. For example, our Head of Solution Architecture Kai Wegner, who can look back on over 10 years in the insurance industry in the areas of process, solution and application architecture. He has served as lead enterprise architect for digitization and board level transformation at one of Europe’s largest insurance companies.

Our VP Product Thomas Hauschild also has more than a decade of experience in B2B enterprise software in highly regulated industries; just to name two of our colleagues.

6) What does the competitive landscape look like? Assuming the market is big and
interesting enough for you to have competitors, who are they and how are you
different?

We know some competitors who also offer AI for document recognition and data extraction.
For example gini in the USA or Zeitgold. What sets us apart above all, however, is our clear focus on the insurance industry. This is what our AI is trained for and this is what our products are geared to. We are not familiar with this from any other provider.

7) You’ve recently received the ISO 27001 certification – what does that mean in
practice?

This means that we deal responsibly with the data that our customers make available to us. This is important for insurance data in particular, as it is often extremely sensitive information. That’s why we decided to undergo the complex certification process.

This involves checking at all levels whether the data is secure, i.e. not only during evaluation, but also in all relevant internal processes. The requirements there are quite high, but we were happy to meet them, because data security is indispensable in the industry.

8) There is a consistent level of fear that AI may end up decimating jobs through automation – what do you think its effect will be relative to the insurance industry in particular?

AI undoubtedly has the potential to fundamentally change work processes, especially in the insurance world. This is already a revolution, comparable to industrialization. And with every revolution, jobs are also lost because they are simply no longer needed. In the end, however, every revolution has produced more work than it has cost.

In other words, yes – certain jobs, especially simple and repetitive ones, will be eliminated. For this reason, numerous new jobs will be created in areas where we sometimes cannot even imagine that they will exist today.

9) On a broader level, what are your impressions on the Berlin tech scene, and how you think it will evolve in the future? Has it been a boon for you to be based there, and why?

It was certainly a blessing for us. From our point of view, there are very few places in Europe where you can build a company like ours. Berlin was at the top of our list from the very beginning.

Three aspects speak for the city: First, it is relatively easy to attract international
employees to Berlin. This is a key point, especially when it comes to developing cutting-edge technology.

Secondly, in the early days, when we were still a start-up ourselves, networking with other young companies was important to us. We learned a lot from looking outwards.

Thirdly, Berlin has an international network of investors and a good set of rules for investing in young companies. And this can also be seen in the tech scene as a whole. Here, many ideas meet great know-how.

Thanks, Sofie!