Skip to main content

Why the future insurance marketplace is data-first

The transformation envisioned by Blueprint Two is only possible if complete, accurate and timely data can support and connect our digital processes.

The Core Data Record (CDR) and the Market Reform Contract v3 (MRC v3) will play a key role in embedding data standards and improving the quality of our data, enabling accurate, standardised data to flow through the entire transaction lifecycle with minimal human intervention. Coupled with the introduction of new core digital solutions, this will transform how accounting and settlement – for both premium and claims – is delivered in the market.

There are already a number of market participants who are using data in innovative ways to help drive innovation and efficiencies.

Alan Tua image

Alan Tua, Director of Algorithmic Underwriting at Ki, discusses how they are harnessing the power of data to drive new technologies and processes. Reach out to if you would like to be part of our data blog series.

Ki is a digital-first, follow-only syndicate operating in the Lloyd’s market - giving broking partners the ability to source capacity online in a matter of minutes. It’s powered by a data-centric, algorithmic approach to follow underwriting and portfolio management which lies at the heart of Ki’s drive to create a digital future in the Lloyd’s marketplace. Originally launched in late 2020 in partnership with Google and UCL, the Ki Algorithm team now provide the technology, vision and execution to deliver a product that generates quotes digitally.

Combining our deep insurance and market expertise with a data-driven mindset from across other industries has helped shape how we think about underwriting in the digital space, and the opportunities available to the wider insurance market.

Core concepts 

To manage the complexities of designing the Ki Algorithm, we employ four core concepts or principles. These are not hard rules, nor are they comprehensive, but instead provide a framework within which we can exploit the technology and the data available to shape our product roadmap. This blog outlines each of them and talks through examples of how they are applied within Ki.


Ultimately, we strive to be as data-driven as possible – eliminating the reliance on intuition, driving consistency and reducing bias. This is best demonstrated in our risk selection, where we have leveraged both Brit and Ki’s data asset to train machine learning models to better assess risk quality at the point of quote. More data means better models, so naturally we do not rely exclusively on our datasets, but look to partner with external vendors to supplement our view of any given risk.

Credibility drives flexibility

We are sensitive to the fact that data availability varies heavily across classes and that emerging risks are not fully captured in historical data assets. We therefore have the ability to constrain or relax the algorithm’s decision-making flexibility based on our confidence in our data and technology. For example, there is more limited scope of line size variation by the algorithm on new business where we have less data available at the point of quote. As we ingest more data from brokers or third-party data vendors we can consider how to safely grow our new business lines.

Controlled by design

One of the core things enabled by a digital-first business model is our ability to execute increased portfolio control when generating quotes. At any point in time the Ki algorithm has full access to data, both within the written book as well as any quotes which have not been bound. This enhanced visibility allows us to control exposure at scale, both within an individual class, and across classes, in a way which would be hard to execute in a more traditional setup.

This control extends beyond portfolio management and into governance and risk management, with Ki able to guarantee which data points are considered when writing any given class of business. Every risk is priced, and all compliance checks are performed at the point of quote.

Human underwriting expertise

Ki was set up as a blend of industry and technology expertise, and this extends to how we underwrite today. 

This is critical to help manage the challenge of data availability within certain classes of business, as well as the human aspect of our relationships with broking partners. The Ki algorithm is one component within our business model, working alongside our Portfolio Underwriting and Portfolio Management functions. We have invested in monitoring capability and flexible interfaces that allow these teams to control our portfolio in a focused way, meaning we can scale our operations digitally and securely, with the right level of human input. 

How each of these concepts get implemented and complement each other is a challenge, but it’s clear that harnessing the power of data to drive new technologies and processes presents huge opportunities and efficiencies for the insurance market. 

Ki’s algorithmic approach is allowing follow underwriting to scale safely, paving the way for a more efficiently transacted market and better service for broking partners and their clients. 

If you are interested in hearing more about what Ki are doing, please visit:

If you would like to find out more on the solutions we are building in Blueprint Two to help embed data standards and strengthen data quality visit:

To give your views on how data can improve the future of the Lloyd’s market please reach out below.

Future at Lloyd's

22 Aug 2023