Why US program business needs AI that fixes the workflow, not just the front end

Paul Templar, CEO, VIPR Solutions

The rapid expansion of artificial intelligence (AI) solutions in insurance is already changing the way carriers and intermediaries conduct business, but for many companies, AI is simply being deployed as a front-end solution, such as a chatbot, to facilitate client interactions.

In the US program market, the challenge companies need to overcome with the aid of technology is more fundamental than a nice-to-have front-end. The market faces continuing operational inefficiencies as businesses attempt to manage program business at scale.

Quite simply, program business is still overly reliant on slow and inaccurate manual processes, where risk, premium and claims data remains fragmented across submissions, spreadsheets, disconnected systems and reporting cycles.

A new direction of travel

A recently announced collaboration between VIPR and WCL has indicated one direction of travel for deploying AI in the program business space, with a view to eliminating the longstanding manual handoff between initial broker submission and downstream bordereaux processing.

However, to understand why this is a potentially significant development, you have to examine the underlying problem in the US program market.

Significant growth in US program business has increased the need for faster onboarding of submissions data, better data quality, cleaner reporting and stronger oversight between carriers, MGAs, brokers and reinsurers.

But this growth and increased demand is at odds with the fact that program business workflows have a longstanding problem with fragmented processes. When brokers send placement messages with bordereaux attached, front-end systems present a top-level view of the risk, but bordereaux data then typically drops out of the digital workflow after initial submission and has to be manually reloaded into downstream systems.

Closing the Gap

Carriers, MGAs and reinsurers tolerate this state of affairs because, for a long time, no single provider has managed to close the gap.

Alongside this, businesses in the program market are looking for solutions to enable easier access to risk, premium and claims data, in order to better inform pricing, reserving, monitoring and reporting on books of business. Typically, the common solution is an unsatisfactory and time-consuming manual workaround.

Unsurprisingly, therefore, the program market is looking for technological solutions to streamline workflows and extract more value from submission data.

With widespread hype about AI’s potential to transform business across multiple industries, there is a pressing need to manage expectations about the reality of what AI can and can’t do in the underwriting space. 

Quality Data Foundations

For program business, the value added by AI is only proportional to the quality of the data foundations that underpin it. Improving program business workflows requires improvements in the processes that deliver those workflows.

It’s in this area that AI-powered solutions are already delivering value, when applied to defined operational problems in the program market: extracting and validating bordereaux data, identifying inconsistencies, improving transaction-level visibility, reducing manual rekeying and helping firms monitor portfolio performance.

Our expectation is that, over the next 12–24 months, we will see a clear separation between firms that can scale effectively by using AI to solve real workflow problems and close the gap between submission and decision-ready output, compared with those whose growth is stalled by using AI as a marketing layer on top of broken processes.