Insurance
Correspondence Automation
In your company masses of correspondences have to be processed and allocated every day. Each mail must be analysed at great expense before the next step can be taken. It is quite possible to assign certain categories, but the problem is that similar facts are formulated very differently, which apparently makes it impossible to automate this process.
-
With semantha® you can automatically classify and assign incoming letters, no matter how they are formulated, regardless of the choice of words.
-
With our data extraction, you can easily extract relevant data points from Correspondences and process them further. The combination with semantha’s® natural language understanding is particularly helpful, as it allows you to find and extract data points in specific contexts.
Document Analysis & Comparison
In your company, documents have to be reviewed regularly and with a high manual effort. In most cases, the content to be reviewed is similar. The problem is that every document is worded and structured differently, which makes automation impossible.
-
In our library, you can define content for which new incoming documents are to be analysed against. semantha® finds the passages based on their meaning, regardless of the choice of words and without any “training” with your data. Thus, relevant contents of a document are found quickly and efficiently.
-
Using our Compare, you can compare already reviewed documents with new incoming documents. semantha® shows you which content matches and which content is missing – out of the box and without time-consuming training with your data.
-
In addition, you can use our Compare feature to compare different versions of a document. semantha® shows you which contents have remained the same and which areas have been changed or removed completely.
-
With our data extraction, you can easily extract relevant data points from documents and process them further. The combination with semantha’s® natural language understanding is particularly helpful, as it allows you to find and extract data points in specific contexts.
Risk Analysis Underwriting
Product managers and underwriters have to check large amounts of text on a regular basis with a high manual effort. Special broker wordings in particular contain many special agreements and are usually formulated individually. New incoming documents have to be checked in detail for undesirable risks under high time pressure, and in the case of new specifications by reinsurers, entire portfolios are analysed for certain contents and exclusions in order to avoid undesirable major claims.
And this is exactly where our AI semantha® can help: It understands text based on meaning, regardless of word choice – so you find content you’re looking for, no matter how it’s worded – out of the box and without extensive training with your data.
According to our experience so far, semantha® saves you 45% of the time when checking risks. Prospective customers thus receive their offers very quickly and your experts can concentrate on value-adding activities such as the acquisition of new customers through individual concepts.
-
With semantha®, incoming documents can be automatically checked for relevant content. These are displayed to the processor regardless of the wording, completely without time-consuming and recurring searches. The comparison with guidelines, exclusions of the reinsurer and already accepted/rejected risks becomes clear and efficient.
Possible applications:
Broker-Wordings, Side Letters, Special risks, Industry framework agreements
-
The Compare function of semantha® can compare the contents of documents with each other, regardless of the wording. Similar passages are compared and marked in color. Areas that were not found in the reference document are also displayed.
For example, wordings to be checked can be compared with your own standard conditions and significant deviations can be detected quickly and easily.
Possible areas of application:
Broker wordings, Framework agreements, Special risks, Competitor monitoring/market comparison
-
Framework agreements in broker sales
A framework agreement proposed by you with the broker can be compared with existing conditions of the competitor.Additional Use Cases:
Analysis of major risks by extracting decisive values from survey reports; Examination of certain contents of existing contracts and agreements in the portfolio (changes by reinsurers, etc.)
-
With our data extraction, you can easily extract relevant data points from documents and process them further. The combination with semantha’s® natural language understanding is particularly helpful, as it allows you to find and extract data points in specific contexts.
These use cases might also interest you…
Consolidation of documents
Harmonization of company-wide regulations, such as company agreements, work instructions & SOP’s across different locations
CV Matching
With semantha® we have developed an AI tool that overcomes the weaknesses of conventional matching tools. The application does not compare whether certain terms occur in documents, but reads the documents at the level of meaning. And it does so within a few moments.
- Ranking of top candidates at the touch of a button
- Overview of the best internal candidates for any skills
- More objective selection of candidates
- Acceleration of the selection process by up to 50%
- More time for interviews to ensure human fit
Knowledge Management
Your company has large masses of unstructured text documents. One of the daily challenges is to search these documents for specific content, whereby normal search functions only offer a simple “keyword search” that does not lead to the desired results. This is particularly due to the fact that, on the one hand, it is not known exactly how a topic was formulated and, on the other hand, the same topics are also formulated quite differently in the various documents.
Bulk Claim Analysis & Comparison
In your company, many procedures on the same topic are processed. Despite always having similar presentations in the pleas and repetitive correspondence with companies, insurers and clients, these documents have to be read and analysed at great expense. Although content can be divided into certain categories, the problem is that similar issues are formulated in completely different ways, which makes it seem impossible to automate the processes.
You have to reply to numerous pleas, such as lawsuits or statements of defence, and repeatedly come across similar submissions. In most cases, you can use excerpts from past responses as a basis for the new plea. However, the classification of the individual submissions of the newly received pleas is the biggest challenge.
Contract Analysis & Comparison
In your company, contracts have to be reviewed regularly and with a high manual effort. In most cases, the content to be reviewed is similar. The problem is that every contract is worded and structured differently, which makes automation impossible.
