Munich Re HealthTech improves insurance analytics with Oracle AI solutions

Munich Re HealthTech (MRHT) aims to solve the data problems faced by actuaries working in the complex world of health insurance. Actuaries use high volumes of often unstructured data from many sources to assess the financial strategies used by insurance companies. For example, the process some actuaries used to gather data to assess financial reserves was effective yet laborious, taking 10 to 15 days nearly every month to make the calculations. The company’s SMAART (Sophisticated Monitoring and Assessment of Risk Tool) application helped insurers bring that time down to just 20 minutes by supporting underwriting, product pricing, claims handling, finance, and risk management in an all-in-one solution.

But as demand grew for SMAART across multiple countries, the MRHT team needed to address customers’ data residency and disaster recovery concerns. Insurers were also asking for new ways to analyze rapidly growing data volumes and sources to draw out business insights. That left the MRHT team in need of ways to develop new reporting functionality faster and use automation and artificial intelligence for more advanced analytics.

By moving SMAART to Oracle Database 23ai on Oracle Autonomous Database and using Oracle Cloud Infrastructure Generative AI, the company improved its security, compliance, scalability, and analytics capabilities.

As a longtime Oracle Database customer, the company chose to migrate its SMAART app to Oracle Autonomous Database 23ai on Oracle Cloud Infrastructure (OCI).  Autonomous Database 23ai provides native AI capabilities such as handling a range of data formats,storing and searching vectors, natural language processing, and integrating large language models. The autonomous cloud database service also provides low-latency data access, top performance, and the flexibility to scale capacity up or down based on demand.

MRHT evaluated several other vendors, whose approach included building a globally distributed database, which would have taken years to migrate the data model and modify the code. Migrating the company’s Oracle Database to OCI and using Oracle Globally Distributed Autonomous Database required no code modifications and delivered valuable built-in database options that improve data security and reporting.

Sharding in Globally Distributed Autonomous Database helps MRHT address important data residency and disaster recovery requirements for its customers. MRHT gains the benefits of using a single logical database while still storing segments of data physically close to customers in multiple OCI regional data centers. In addition to the data residency and disaster recovery benefits, that proximity helps deliver low latency and faster reporting times.

OCI Generative AI, a fully managed large language model service, and Oracle APEX, a low-code development framework that automates app development and data integration, contributed to the company’s choice.

Moving the SMAART app to OCI helped MRHT achieve numerous goals, including data residency compliance. Many countries are implementing data residency requirements that come with significant fines for noncompliance. GloballyDistributed Autonomous Database gave MRHT an economical way to meet those requirements, with minimum support and maintenance required. Built-in, automated database sharding policies allowed the company’s customers to store segments of data in the OCI cloud region in the country of their choice, helping them improve data access and control by allowing a single logical database to be distributed over multiple geographies while keeping the application in one region. Data proximity and distributed queries also reduced global data replication by eliminating the need to copy large amounts of data to and from a central warehouse, resulting in faster reporting and analytics. Globally Distributed Autonomous Database also helped lower data security risk by restricting user access to specific authorized areas.

Oracle APEX, a low-code application development framework that’s included with Globally Distributed Autonomous Database, let the company quickly create an automated means of extracting data from spreadsheets, reducing the time it takes to build analytical dashboards in SMAART from 15 days to about 20 minutes while also reducing human error. Within months of starting to use Oracle APEX, the MRHT team was able to create the dashboards on time, on budget, and without the need for additional software licenses.

MRHT also built an AI chatbot as a companion for actuaries, underwriters, and senior management so they could use SMAART to search its knowledgebase for documentation and advanced data analysis. Using natural conversational text, the chatbot delivers instant answers and insights in multiple languages. Built with OCI Generative AI and Oracle Autonomous Database 23ai’s Select AI, AI Vector Search, and APEX, the AI chatbot accurately answers more than 90% of the questions posed to it. The tool helped accelerate data access, reduce IT dependency, and improve decision-making. In addition, the company saw lower custom development costs, better portfolio management, and stronger compliance and gained real-time insights for optimized pricing, risk assessment, and underwriting strategies globally.

Oracle partner and APEX expert Pretius assisted MRHT with this project and provided guidance on the ongoing management of APEX applications. The project went live in four months.