Bitesize InsurTech: MarkLogic
September 28, 2018 Greg Brown
MarkLogic is a database that sits on top of legacy systems to create a unified and actionable view of data.
The product allows (re)insurers and brokers to pool information from unstructured data sources such as emails, PDF records and texts. The database then organises the data in a hierarchical structure in real-time to allow underwriters, for example, to immediately access the information needed to evaluate risks. The database has other applications across the insurance value chain, including claims and back-book optimisation.
Relational vs. non-relational databases explained
Many readers of this blog will have heard techies discuss the relative merits of ‘relational’ and ‘non-relational’ databases over the years – but been too afraid to ask what the difference is. Today is the day that the fog lifts.
Most insurance legacy systems use only relational databases. These databases store data in tabular formats, much like Excel. They are sometimes called SQL databases.
Relational databases are helpful for storing and comparing structured data such as area codes. They are less effective at combining unstructured data (e.g. emails), a major challenge for insurers given the shift of focus from structured (e.g. basic proposal form data) to unstructured data (e.g. data augmentation in the underwriting process).
Non-relational databases, sometimes referred to as a NoSQL databases, are the ‘next generation’ of databases and allow for larger and more complex datasets to be stored and analysed. Their main uniting characteristic is that they are not relational.
MarkLogic provides a non-relational database to allow insurers to combine and analyse structured and unstructured data from around the business.
The company began in 2001 and introduced its product to the insurance industry in 2014. Its CEO Gary Bloom joined in 2012, having previously served as Executive VP at Oracle, one of MarkLogic’s major incumbent competitors.
MarkLogic claims it can deliver production-ready systems at a fraction of the cost and time of other database transformations (3-6 months) since it does not require insurers to abandon their legacy systems. Rather, their database works alongside existing systems to provide a more optimal balance between relational and non-relational data storage.
MarkLogic recently worked with Erie Insurance to improve claims processing. Instead of having to switch between data sources and programs when processing claims, the claims team were able to access live customer information in one location, greatly reducing claims processing costs and time.
The Oxbow Partners view
Data is a major theme in the insurance industry. In our recent report for the Lloyd’s Market Association on the impact of InsurTech on the Lloyd’s market, data was an enabler for all five trends. Whilst data creates clear value for insurers (for example, greater risk insight in underwriting and analysis of customer experience) the challenges of accessing data from legacy systems prevent many from realising this value. Given the cost of replatforming, MarkLogic’s solution, which allows insurers to remain on legacy systems, is attractive.
The challenge for MarkLogic will be getting noticed in a market dominated by bigger vendors such as IBM and Oracle. Additionally, the direct buyers will be CIOs but the value will be delivered to CUOs and COOs creating an additional proposition communication challenge.
One way that MarkLogic is solving this problem is by being ‘use-case led’ rather than ‘technology-led’. We think that this will make them more relevant to our clients.