At many financial companies we observe problems with calculation engines. For instance the actuarial calculation engines. In this case the actuary knows exactly how the calculation rules are meant to be, but the long chain between business and IT leads to suboptimal usage of this knowledge. This causes several problems.

Long lead time
It takes a lot of lead time before a calculation engine gets a production status. Changes in specifications take a lot of time to be adjusted in the calculation engine.
The calculation engine contains faults and bugs which could have been spotted quickly by domain experts. This is caused by the large distance between the domain experts and the developers who implement the calculation engine.
Dependencies on the supplier
Domain experts are dependent on external suppliers to make adjustments to the calculation rules.
High testing effort
Testing calculation engines is a difficult and (time) expensive process. Testers are commonly duplicating the development effort by making their own calculation engine in for instance Excel.
High costs
The total process is too expensive due to the many competences needed in the whole process and rework of the supplier.

Our solution

Our calculation engine is a good example how model based working can help. We see that the problems mentioned above are caused because the domain expert, in this case, the actuary, is too far in the chain from the final product. They are extremely well educated and able to make a calculation engine on their own. They commonly make this kind of engine already in for instance Excel. Their expertise can be used much better.

Our platform enables domain expert to specify, document, test and eventually generate working software themselves. This is made possible by or model based working approach. Domain experts describe the calculation rules in a model using our domain specific language. Actuary can master this language after a training of only 3 days. The resulting models are sufficient for our platform to automate the rest of the process.

This approach has many advantages.

Automatic control on specifications
Our platform checks the modeled calculation rules on correctness and executability.
Generate correct documentation
The calculation rules can be exported as HTML, Word or PDF documentation to be shared within the organization.
Automated testing of the calculation rules
Our platform enables automatic generation, execution and evaluation of test cases to test any calculation engine against the modeled calculation rules.
Generate working software
As the calculation rules are modeled unambiguously, our platform can generate a well performing calculation engine with web service interface.

The calculation engine at customers

With our approach we have implemented calculation engines at several customers. For instance at Achmea, a big insurance company in the Netherlands. Our customers are always satisfied and they experience the advantage of our progressive approach.

Testing of our calculation engine would have cost thousands of hours and was not possible within the deadline. With the Axini platform, a thorough test costs only several hours.


Is the calculation engine a solution for your organization's problems? We would love to tell you more. Contact Machiel van der Bijl for additional information.