Milliman is enabling self-service actuarial modeling with Microsoft Fabric

Insurance companies constantly undertake and mitigate risks, so that even in a downturn or disaster, they remain solvent. Any calamity has the potential to dramatically change the economic equation of an insurer. Actuaries are the experts who use probability, mathematics—and large amounts of data—to understand, predict, and forecast risk. Insurers depend on their savvy and skills to keep their businesses operational.

Milliman is a worldwide financial services company that works with its insurance industry clients to manage emerging risks and advance financial security, so millions of people can live for today and plan for tomorrow with confidence. Managing any risk is complicated, but life insurance can be particularly challenging due to contract durations that can span decades. With such long-lasting business, many insurers rely on legacy systems that don’t necessarily connect with modern infrastructure. Insurers have faced difficulty building an enterprise data warehouse because it needs to be populated from such a plethora of legacy systems, which is both expensive and increasingly cumbersome as the business evolves. 

“The data necessary to model and understand our business is complex and requires a deep understanding of the business to process. We have found that empowering actuaries to work closely with the data to build models speeds up delivery and improves quality,” says Tom Peplow, Principal and Senior Director of Technology Strategy for Life Technology Solutions at Milliman. 

Milliman needed to empower its actuaries, and the actuaries of its clients, with the ability to better address business needs through a modern data analytics solution. Peplow and his team were required to combine complicated data platform components to create a bespoke solution that allowed their actuaries to build custom data pipelines for their clients.

“These data manipulations are so complicated that it’s difficult to explain to IT teams how to help resolve them on behalf of the actuaries,” Peplow explains. “In addition, actuarial teams provide important information to regulatory bodies and analysts, so they need tools to analyze terabytes of data to be highly confident in the results quickly.” Milliman already used Microsoft Power BI for analytics and reporting and wanted to ensure it was integrated into any future analytics solution.

Implementing Microsoft Fabric, thoughtfully, step by step

Enter an all-in-one analytics solution for enterprises that offers a comprehensive suite of services including data lake, data engineering, and data integration. It’s no longer necessary to piece together different services from multiple vendors. Microsoft Fabric provides an end-to-end, easy-to-use product that brings together new and existing components from Power BI, Azure Synapse, and Azure Data Factory into a single integrated experience. 

Microsoft Fabric helps businesses accelerate data value by centralizing all data and analytics workloads on a single, open, governed foundation. This allows users to build artificial intelligence (AI), analytics applications, and machine learning (ML) models on a data lake-based architecture quickly and easily. Additionally, Microsoft Fabric empowers organizations to expedite innovation with secure access to self-service analytics tools that facilitate near-real-time insights. 

“Microsoft Fabric would have been our choice from the beginning—had it only existed then,” says Peplow. Milliman has been giving users access to actuarial tools so they can unlock data and produce analytics without involving IT. However, the actuarial domain is very complex. The team had to build their own  data platform, distracting the team from the key business problems while increasing long term total cost of ownership. “Building a democratized data platform was just too far beyond our specialization in actuarial software,” he recalls. 

A unified infrastructure, plus the ability to keep Power BI front and center, led Milliman to take a considered, step-by-step approach to deploying Microsoft Fabric. Early results have been very encouraging.

Microsoft Fabric directly addresses challenges previously faced by both Milliman and its clients. For example, the company was offering Microsoft Power BI embedded, but using the OneLake component in Microsoft Fabric makes data available faster, expediting data access. OneLake also eliminates direct query limitations and provides import mode performance for easier, more efficient reporting. Users can now define data permissions that are honored on all the tools they use—something not previously available. And that’s just the beginning.

As actuarial modeling may require access to personal data and company financial data, Milliman systems must be secure. Milliman needed to find the appropriate balance between data security and data access so people could do their jobs. Microsoft Fabric has helped Milliman navigate these constraints, enabling the right people access to the data they need without sacrificing data integrity or data security. 

Many actuaries have historically experienced limits to the analysis they can perform. With Microsoft Fabric, embedding notebooks along with Power BI enables more advanced analytics. Power BI artifacts are version controlled in GitHub, enabling integration with existing Milliman actuarial models and process logic code.

There will also be notable benefits for Milliman internally. For example, moving data between internal services can be time consuming. Integrating all services around a single data lake within the OneLake storage system of Microsoft Fabric means fewer copies of information, thus reducing latency and cost. Milliman provides a custom data gateway for moving data between customers’ storage and its environments in Azure. Milliman can federate its data with customers via OneLake and leverage a single security model. 

Microsoft Fabric can also help the bottom line. For example, users pay just once for shared capacity across workloads, simplifying management and reducing costs. Azure Data Factory mounting provides a path to continue to produce data in existing data pipelines but benefits from new capabilities, meaning Milliman does not have to undertake an expensive migration of complex processes.                              

Although Milliman is still discovering how Microsoft Fabric can benefit both employees and clients, Peplow is already thinking about the future. “Microsoft Fabric has the potential to be the vehicle for Milliman’s data transformation and data platform that we can confidently market to our clients,” he says.

Ken Mungan, Milliman Chairman, added, “Milliman is tremendously excited about Microsoft Fabric. We see that Microsoft Fabric will be critical for our professionals and insurance industry clients to realize the full potential of AI to protect the health and financial well-being of people everywhere.”

Modernizing actuarial modeling with AI

A challenge for Milliman has been that all data science and AI use cases were managed outside of its proprietary Integrate platform. Now, there’s a solution. “Microsoft Fabric is the future of Microsoft’s data vision. AI needs data, so Fabric needs AI,” Peplow says. “The bet is investments will continue to converge and Microsoft Fabric will make it simpler to realize opportunities from AI. Fabric artifact embedding, enabling our Integrate platform to be built on Microsoft Fabric, can support our product’s advancement while streamlining our infrastructure.”

And as every actuary knows, you can’t argue with data. “Actuarial modeling is a complicated problem that exists across all parts of the insurance world,” Peplow says. “With the help of Microsoft Fabric, we will solve that problem specifically for life insurance companies.”   

Find out more about Milliman on Twitter and LinkedIn.