Enlighten your straight-through processing. With a solution that understands your business mail better than any before.
As full process or single modules
From OCR and text import, to classification, extraction and validation
Expandable, trainable, and actively learning
Standard recogniser extendable to any specification, to any language, across the entire process chain
On-premises and highly secure
100% compatible with your existing system, fully compliant with data protection requirements
How insurance companies benefit from it.
90% of the text of a damage report is in the vast majority of cases not important for your agents. Our software extracts only the relevant information, and saves your employees valuable time.
ICD-10 is one of the most widely used taxonomies of diagnoses and has 50,000 entries. While a person cannot remember all of these, our machine knows them all. It automatically assigns ICD-10 codes to the diagnoses in your medical documents.
Certificates of incapacity
Which insured person was on sick leave from when to when? Thanks to the Cognitive Workbench, this information enters your system entirely without manual input.
We cannot prevent cancellations, but we can make the process efficient. If specified, our system can also identify the reason for termination. Perhaps it can even be used to change the customer’s mind.
Who wants to have paid for what and why by when? Our machine will tell you. If you want, even down to the individual invoice item.
Notification of claims
In the event of damage, fast processing and assistance are what count. Don’t waste any time. Let our machine bring the most important information to your systems quickly and automatically.
It is always cheaper to retain existing customers than to acquire new ones. We also recognise complaints by their tone of voice and can direct them directly towards processing.
It works differently
It examines documents differently
Instead of just capturing text, data or images only, our system can process them all together. It learns not only what is in the document, but also where it is placed. It uses this spatial context to recognize meaningful connections – and thus extracts more useful content.
It detects correlations and causations
Rather than looking at a word or a number in isolation, our machine also analyzes what happens to the left of it, to the right of it, above and below it. This multidimensional recognition ensures a much more precise correlative and causal allocation of information. This is essential, for example, in the automatic recognition of lists and tables in documents as expert reports or invoices.
It boosts your learning curve
Rather than depending on any service provider to adapt your software for you at some point, you simply optimize our system in-house. Define for yourself what our machine is supposed to recognize and train it without delay. The result is a system that is constantly evolving – offering you the chance to gain decisive competitive advantages.
Train fast. Apply immediately.
In a closed data loop.
The Cognitive Workbench: Every newly trained skill can be applied immediately and productively. With just a few clicks you can create a docker container with selected recognisers. You can then integrate it directly into your internal systems. Your systems in turn deliver feedback right back to the Workbench, from which it learns again. Scalable as desired. Flexible control. On-premises.
Understands every language. Learns any knowledge.
Trains without a break.
Our Cognitive Workbench is not a rigid, rule-based software. It is a true Enterprise AI platform: a universal, intelligently learning text and document mining system. It can be trained and configured highly specifically for any language, any knowledge and any application area. Simple and efficient. And again and again if required.