Sponsored by: ?

This article was paid for by a contributing third party.

Blog: Driving efficiency and automation by breaking down data silos


  • The London Market controls more than $91bn in gross written premiums
  • Increasingly speciality insurers are looking for ways to automate the underwriting process
  • Database technologies can offer an agile and flexible solution

Some underwriters are still reviewing data manually but Rodrigo DeCossio, UK insurance lead at Mark Logic, suggests 2018 might be the year to automate these processes and drive more efficiency in underwriting.

Conventionally viewed as the leading global hub for insuring commercial and speciality risk, the London Market controls more than $91bn in gross written premiums.

Within the market, there is clearly a huge amount of segmentation and with this come many different classes of risk. This is even more the case within speciality insurance in areas such as aviation and marine, which typically require more bespoke policies.

However, the more specialised and complex the risk is, the more data is required to price and underwrite it. Consider the aviation industry for example. An underwriter will need to review information from multiple sources, which are generally document-based, and could include maintenance schedules, flight paths and aircraft specifications. This process is typically a manual one, done by hand to extract all the relevant data from documents that may include a combination of Excel, Word, PDF, images and emails, and aggregating it to gain context and a better understanding of the risk.

But increasingly speciality insurers are looking for ways to automate the underwriting process by extracting all the relevant data from various silos and combining it with external sources. In this respect, the speciality insurance market faces arguably its biggest challenge as one of the last industries yet to be truly disrupted by the digital revolution.

When you reflect that insurance is essentially a business of managing risk, it’s quite amazing to think that insurance companies can still accurately calculate their risk through this manual review process when it’s very difficult to quickly and effectively access all of the data they need.

For example, take the case of a chief business actuary at a major global insurer with $2bn in annual property premiums. They need to view the aggregated property risk the business has in a specific geographical point or defined area, which currently takes actuaries and pricing teams a lot of manual work to find the information needed to calculate.

The challenge is this information could be spread across upwards of 20 data silos, with no efficient way to search across them for the specific location. It’s having this single source of truth across the business that is crucial.

The underwriter will need to manually view these reports, cross-checking them against the guidelines within their policy and then approve the application, turn it down or ask for more information. This whole process can take up to 45 minutes, and insurers will manage thousands of such applications every month. Automating this process would reduce the time down to less than two minutes, allowing the company to handle many more applications.

Different database technologies that enable underwriters, claims handlers, analysts and actuaries to quickly extract, aggregate, link, search and query the data within their organisation offer an agile and flexible solution. With a 360-degree view of data anytime and anywhere, the insurance industry could automate the entire process of information review, risk management and pricing, and ultimately enable better decision-making

Rodrigo DeCossio Mark Logic
Rodrigo DeCossio, UK Insurance Lead, Mark Logic
  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have an Insurance Post account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: