CAI - one solution for a million challenges - Pro Global

CAI – one solution for a million challenges

From turbo charged underwriting performance to more accurate risk modelling.


October 14, 2021

Share this page

From turbo charged underwriting performance to more accurate risk modelling, Sue Barenbrug, Pro Global Head of Data Management and Risk Modelling, explores the applications of CAI in today’s fast paced market

It was British mathematician and entrepreneur Clive Humby who coined the phrase “Data is the new oil” –  a well used trope that was later refined by American physician Michael Palmer, who added that data is “valuable, but if unrefined it cannot really be used”.

It’s the latter nuance that we couldn’t agree with more here at Pro Global – the fact is that unrefined, unstructured data is about as useful as a chocolate teapot. It simply has no practical application at scale.

Addressing everyday data challenges

In developing CAI – our Cleansing Augmented Intelligence platform – we wanted to directly address the data challenges re/insurers face every day when checking and updating data from millions of exposure schedules or Statements of Value (SOVs).

The central challenge is that the market that recognises the value of good quality data, but also recognises the simple truth that highly skilled individuals should not be spending their working days checking repetitive data records.

Learning algorithm

This is where augmented intelligence comes into play. CAI is an augmented software platform, which means it uses a learning algorithm – once a human has shown it how to do something, it doesn’t need to be told again. In fact, it extrapolates this learning across its applications to make every task more efficient.

CAI is currently focused on cleansing the SOV data for land based property risk that feeds your catastrophe and exposure models, transforming the speed and accuracy of the data; the fuel that powers your proprietary modelling outcomes. This powerful tool has the ability to provide multiple outputs for different model input formats from one cleanse, setting the new standard for data cleansing.

Here are three examples of the challenges CAI addresses:

CHALLENGE: We can’t quote fast enough

SOLUTION: CAI’s learning algorithm recognises data previously cleansed and stops scrubbing the same data. CAI’s insights also improve the accuracy of broker submissions, and allow far quicker quotes to be submitted from underwriters. In this way, CAI powers much faster quoting and demonstrably increases the percentage of pre-bind modelling.


CHALLENGE: My data isn’t geocoded accurately and it’s too time consuming to check each record

SOLUTION: CAI is a powerful tool that allows all locations to be 100% geo-coded, to provide better data quality, powered by augmented verification tools that allow for clear data checking to dramatically reduce the risk of error. CAI’s improved geocoding accuracy and detail also powers greater risk modelling accuracy and speed of insight – in turn powering truly data-driven underwriting decisions.


CHALLENGE: Our skilled staff are being bogged down in mundane work

SOLUTION: CAI can cleanse huge volumes of data in one go, only directing human resources towards processing anomalies or new information previously unseen by the system. This means that your team’s skills are only directed towards the true problem solving demands and frees up time to focus on more value-added activities such as better client service, improving the capacity and profitability of your team.

Just like oil, data needs to be refined before it can power a business to its full potential. And the reality is that there are millions of exposure schedules and SOVs that are cleansed and re-cleansed each year in our industry – representing a wealth of unrefined data ripe for exploration.

Get in touch to find out more about CAI, or request a demo.

Unlock Your Treasure Chest:

Unlock Your Treasure Chest: Navigating the Complex World of Loss Fund Repatriation in the London Insurance Market In our recent article that my colleague Chris […]