AI in underwriting
AI technology does not replace an underwriter - but it can significantly accelerate processes. Using Natural Language Processing (NLP) with Deep Learning methods, kinisto extracts important data and core information from documents of any kind and displays or highlights them in a structured way - regardless of form, structure and wording. This means kinisto can extract data from doctors’ letters - for process automation in advance risk inquiries. kinisto also recognizes possible risk clauses in contracts and thereby creates the basis for significant process acceleration.
Automation in Underwriting
kinisto recognizes information in context and makes it usable. For the verification of insurance applications or preliminary risk inquiries, this means that data from important accompanying documents can be captured in a structured manner - regardless of the layout and form of the documents. In addition, kinisto can perform automated checks for specific features that may be associated with risks.
Classification & data extraction
kinisto recognizes all text elements contained in the document and puts their connotation and position into context with each other. This way, relevant information can be extracted even from non-standardized, long or complex documents.
Process optimization
The extracted data enables accelerated application review and can also be used for automation of downstream processes.
Automation of standard processes
kinisto supports underwriters in their daily work by quickly and accurately reading out standard information and standard clauses. Important information no longer has to be searched for “manually” in documents. kinisto can immediately output recurring relevant information, thus facilitating repetitive tasks and massively accelerating the entire process. kinisto can also highlight conditions that deviate from the standard or critical formulations and thus support risk assessment.
Preliminary risk inquiries - private health & occupational disability insurance
kinisto extracts relevant data and information from doctors’ letters - even if they are submitted in a wide variety of formats. This allows applications to be processed faster and in a more structured manner. At the same time, the extraction of insurance-relevant information enables risk assessment to be automated to a large extent. The data can serve as the basis for a traffic light system, for example.
Industrial insurance contracts
Quick overview: kinisto can highlight all important information in non-standard contracts, even with heterogeneous and divergent wording - including information on risks such as coverage commitments incl. coverage amounts and sub-coverage amounts. This means that even pre-formulated applications, which have initially been written by brokers, can swiftly be processed and checked for possible risks.
Risks in the contract portfolio
kinisto makes it possible to analyze large volumes of text and documents. Thus, the entire contract portfolio can be checked for specific risks. Coverage modules and sublimits in the contract portfolio can be read out from large quantities of contracts. The information can be used, for example, to quantify the exposed risk at the sub-cover level.
Profitability measurement
kinisto can read data from claims reports. This data can be linked to the contractually defined coverages. Thus, it is possible to quantify the profitability of individual coverage components - the information can be a valuable basis for further business decisions.
