Solution Architecture:
We can work with you to help design an overall solution around the R³S Suite according to your IT policies whilst maintaining compliance. The following diagram displays an architectural overview of RNA Analytics software and services
Compute hardware:
The hardware used to run models can have a significant effect on their performance, so scaling the hardware to the model requirements is critical. Our actuarial consulting service will look at your models and inform our technical services to identify the CPU core and RAM requirements to support optimal calculation performance. This is usually carried out in conjunction with the Model profiling for performance service as it can significantly impact the optimal hardware required by your models.
Data storage:
The volumes of data required to run models and output generated from them is increasing significantly as companies need to understand and analyse their results. The complexity of the data needing to be stored is also increasing with changes in regulations, such as IFRS 17, defining information and results that need to be maintained at more granular levels than previously. We can work with you to estimate target storage requirements and help architect an optimal data structure strategy, including reviewing your existing storage solution and making recommended changes if required.
Integration:
A diverse range of source data and regulation driven results can necessitate multiple integrations. RNA Analytics has the expertise to build, or advise on building, new ETL processes as required by solution architectures for a wide variety of source/destination file and database formats. In addition, we can review R³S models to potentially optimise external transformations as well as assess the suitably of current in-place processes for a target solution.
Audit, Control and Security:
A crucial aspect of any IT environment supporting the production of regulatory calculations is automation, security and traceability. We can work with you to eliminate manual intervention in the production process, either using tools such as R³S Process Manager, R³ Toolkit and ETLs or new R³S functionality such as Writing to databases. Those tools can also be secured for authorised roles and audited thereof.
Reporting:
Regulations, business analysis and decision making will likely have different requirements to understand model result data. We can implement or advise on reporting strategies across a range of needs, from interactive analysis driven reports to enterprise wide static report serving, generated either on premise or in the cloud.
Cloud:
As part of the solution design process we can look at the viability of utilising cloud services to support some or all aspects of the target operating environment you wish to implement. This can range from utilising temporary “pay for what you use” compute capability and results held in a cloud-based data warehouse to a full cloud solution, consuming on-premise source data to produce cloud-based reporting.
On-going management & maintenance:
We can help you to put processes in place to make your actuarial modelling environment easier to maintain and manage in the future, both practically and from a cost/budget perspective. Built into a solution design can be estimates for current and likely future resource needs such as projected data storage growth and monitoring of system resource usage with a view to improving hardware resource configuration. Additionally, we can help you to implement processes to manage the impact of new releases/upgrades and model changes, including implementing test environments.
R³S Process Manager:
We can assist to you to get the most out of your investment in R³S Process Manager. In particular, we can help you to implement a model governance framework within R³S Process Manager. Including:
• Designing model governance as part of the overall solution architecture
• Implementing and configuring R³S Process Manager
• Administrative user training and enablement
• Review of current R³S models to set up for use in R³S Process Manager
Model profiling for performance:
This service is aimed at improving the performance of existing models. This can be helpful for example, in ensuring existing cashflow models are fully optimised to support IFRS 17 calculations. There is no guarantee the model efficiency will end in models with quicker runtimes but previous results from other client commissions have resulted in various degrees of efficiency ranging up to 75% improvements in run times with no changes to reported key values.