"The Renaissance insurance" increased profitability of retail business using SAS Real-Time Decision Manager and machine learning

2019: Creation of the central decision making system for preinsurance checks of retail types of insurance

On August 26, 2019 the Renaissance Insurance company reported that it decided to create the central decision making system for preinsurance checks of retail types of insurance and also to implement predictive analytics for increase in accuracy of risks assessment according to requests for agreements. Since spring of 2019 a system works in the mode of trial operation and already paid off.

According to the company, for creation of a system the company considered exclusively large players in the market business analysts. When choosing compliance SOFTWARE to requirements to functionality, productivities and fault tolerances was considered. As a result for implementation of a system selected solutions SAS Real-Time Decision Manager and SAS Enterprise Miner. For benefit of SAS, in addition to compliance to all requirements, told experience of implementations in bank the sector RUSSIAN FEDERATION. One more important advantage SAS is that the command of vendor could act for Renaissance Insurance as the business consultant and was ready to adapt practice of creation of decision making systems on the basis of experience of both the Russian, and foreign projects.

To develop internal examination, decided to lead Renaissance Insurance implementation by own efforts. At the first stage designed decision making system and unrolled software modules. At the second stage the project team connected a system to internal and external data sources. Further configured the strategy of automatic decision-making according to requests for insurance contracts. Strategy combined difficult business logic of segmentation and tariffing with the models of machine learning constructed using SAS Enterprise Miner especially for increase in accuracy of risks assessment according to requests for agreements.

When the decision making system was brought into trial operation, its operational and cost efficiency was proved on a small flow of requests. The term of implementation of all project, taking into account complexity of an IT landscape of the company and need of development of internal competences of profound analytics, took 9 months.

The implemented project is of special value for retail business. First, we accurately see that the unprofitability on a portfolio of retail agreements decreases. It already provided to the company a certain growth of profitability, and in the future will allow us to offer even more attractive terms of insurance for clients. Secondly, at the expense of SAS Real-Time Decision Manager we managed to centralize decision making on preinsurance checks – to make so that key fronts and partner services "went" behind rates to one point. At the expense of it we make changes in the strategy of underwriting and tariffing only once and at once in scales of all company now. And thanks to visual interfaces of SAS we do it independently and quickly, with the minimum participation of other departments. Finally we are going to convert this increased flexibility into growth of rates of business development: to display products in sale quicker, to connect partners quicker. At last, we expect that all designated effects will amplify when we transfer to SAS Real-Time Decision Manager of strategy of decision making on all retail products and also completely we migrate in the created system tariffing. Sergey Demidov, the managing director of department of insurance of vehicles told Renaissance Insurance

Increase in marginality of business, expansion of a market share, accumulation of stability due to development of networks of agents and partners – here things which for August, 2019 are the focus of attention of retail insurers. However we see that on the way to these purposes many insurers have an obstacle – the heavy saved-up technical debt regarding systems for preinsurance check. Separate IT solutions, the long-term not documented code, a difficult situation with quality of data – all this does not allow the companies to solve key problems for achievement of strategic objectives: begin to estimate more precisely risks according to requests for agreements and quicker to make changes to logic of segmentation and tariffing. A way to the solution of these tasks – consolidation of systems for preinsurance check and creation of the central decision making system. Stepan Vanin, the head of business solutions for an insurance sector of SAS Russia/CIS told