Technology
Part I: Overview
The interaction of the algorithms in the calculation kernel with the graph databases enables intelligent digital conversations.
THE CONVERSATION IS IN THE CENTER
Riskine models digital consulting processes for leading insurance companies and banks using graph databases.
Question-answer games (conversations) are highly complex and every answer can and should lead to a new question and not follow a rigid straight-line pattern. Linear questionnaires quickly become a torment for clients and advisors.
Since the connections between the data are also stored in graph databases, particularly efficient and flexible queries are possible. The next question arises from the previous answers and both the questionnaire and the answers are constantly adapted to the client (gender, hard or soft phrases, answer options, language etc.).
INDIVIDUAL
CALCULATIONS
during the process
Our algorithm (riskine core) processes the information obtained from answers simultaneously. The results in turn flow into both the questions and the answers.
Examples are:
- Product recommendation & prioritization (through scorings)
- Proposals for coverage amounts
- Identification of customer risks (risk profile)
- Gap calculations (e.g. pension gap)
- Price calculations
ALL SERVICES
AVAILABLE VIA API
All calculation algorithms and services from our products are also available to you as a single / modular web service. We rely on a modern RESTful architecture with JSON as exchange format.
Our API consists of various services that offer not only the riskine functions known (risk calculations, coverage totals, product prioritization) but also various additional services such as pension calculator, address service, gross net calculator, flood zones, and many more. Our Question-Answer Process Engine (POST /nextq) can also be queried through our Web API.
VARIABLE DESIGNS DEPENDING ON THE APPLICATION
The question-and-answer process engine allows the automated optimization for the respective frontend. Frontend-specific texts can be generated from the conversation model. For example, your customers receive brief questions and answers when interacting via chatbot or long texts when their application is opened on a tablet. The same mechanism applies to different languages.
SMART AND ADAPTIVE TEMPLATES FOR HIGH FLEXIBILITY
The questions from the question-and-answer process engine become intelligent / highly adaptive templates via grouping functions. The next question within the template adapts to the previous answer, adopts information, and performs calculations based on the previous answers.
for mobile applications, websites and client portals
All algorithms and calculations needed for an intelligent consultation are integrated within our chatbots. The dialogues usually include less questions, contain natural language understanding (NLU) elements and interact with the advisors’ solution. Besides providing advice in a normal insurance consultation, the chatbots can be used to answer FAQs or manage claims etc.
See the second part:
TECHNOLOGY:
ADVANTAGES
