Appointed to the Reassured executive committee last year, Beth Whelan (pictured) is steering what she describes as a “strategic reinvention of the business” built on cloud data infrastructure, AI and redesigned digital journeys and underpinned by a deliberate cultural shift towards customer centricity.
She joined Reassured after a long career at Equifax, where she rose from account and product management into a C-suite level role overseeing strategy, change, product, CX, ESG and operations. That experience in data-heavy, highly regulated markets is now being redeployed in life insurance, a sector she freely admits was new to her when she arrived.
“I landed at the perfect time,” she said. The core data platform work had already been done before she arrived, “Our IT director had done all the hard work of putting the data platform in and I got to come in and actually start to do the interesting stuff.”
For Whelan, digital transformation at Reassured is not primarily about channel shift or front-end polish. It is about using data and AI to move from one-off policy sales to long-term relationships.
She characterises the change as a move “away from transactional customer engagement to being much more customer centric”, with the ambition to “anticipate customer needs and adapt to their needs across their lifetime, versus one-off transactions.”
That ambition reflects a growing recognition across the market that customers’ protection needs evolve considerably over time and that many policyholders simply do not adjust their cover accordingly. Whelan points to her own family as a cautionary example. Her twin sister, who lost her husband, discovered they had never updated their life insurance after moving home and faced a shortfall as a result. The reminder email had almost certainly been sent; it simply passed unnoticed at a time of competing priorities.
Reassured is trying to use its 15 years of data to address precisely that kind of gap. The firm has been investing in a cloud-based data and AI platform and “digital journeys that then enable customers, if they choose to, to fully self-serve their life insurance purchases through a digital journey.” The longer-term play is to connect those journeys to richer behavioural, operational and outcome data, so that we can “really start to think differently about that kind of end-to-end customer journey and experience” and “make sure that we are meeting what those customers’ expectations are.”
In contrast to much AI rhetoric in financial services, Reassured’s leadership is emphasising growth rather than contraction.
“We recognize that there are more people who want and need life insurance, and who we’re servicing today,” Whelan said. “So for us, it’s not about creating efficiency so we can be smaller, it’s about creating efficiency so we can help more.”
A more scalable, more efficient operation is therefore framed internally as a way to expand reach and improve service rather than a prelude to job cuts. That framing matters when rolling out automation, particularly in areas such as quality assurance that can look vulnerable when AI is introduced.
To keep technology and change investments aligned, Reassured has defined six “big themes” that anchor its annual strategy. These themes, agreed at executive level, encompass customer centricity, lifetime value, data and AI innovation, digital transformation, back-office automation and people and engagement.
Any new idea or proposal is first tested for “strategic fit” against those big themes. “If it’s not actually touching any of those themes, then it’s an easy one to say, actually, it might be a great idea, but it’s not the focus for this year,” Whelan explained. That discipline is designed to avoid the familiar trap of taking on too much and diluting impact.
Where there is alignment but limited evidence on benefits, the default is to pursue proof-of-concept rather than a fully-funded programme. “We’ll try and then chunk it up, so let’s do a proof of concept. Let’s work out whether we think this could work before we do that full rollout or we do that full investment case.” The intent is to preserve rigour around investment decisions “whilst not stifling creativity or innovation”.
Ideas are no longer only expected to come from the senior leadership tier. An innovation forum has been launched both as an outlet for participants in the firm’s new data and AI academy and as an open channel for operational staff to surface process improvements. The forum includes internal leaders and external subject-matter experts to help shape and assess proposals.
One of the most striking examples of AI deployment to date is within Reassured’s quality assurance operation. Historically, the team was able to audit around 12% of calls. Whelan said the firm can now audit 100% of its two million monthly calls using an AI transcription and analytics tool aligned to a 46-point QA framework.
“It allows us to analyse all of the call against the majority of that framework,” she explaind. The tool scores the framework, flags potential issues and risks, leaving human reviewers to focus on targeted follow-up. For an intermediary, handling high call volumes, the shift in coverage and insight is significant. This is enabling improved feedback to our agents, which in turn is improving customer experience.
Unsurprisingly, this type of automation can trigger fears among staff that “if you’re automating QA on 100% of calls, you won’t need me anymore.” Reassured tried to confront that concern head-on.
Ahead of go-live, the company planned resourcing based on the future implementation, backfilling any shortfalls with support from managers and team leaders. “We really didn’t want to put AI in, gain efficiency and then immediately need to reduce headcount, ” Whelan said. Some team members have moved into growth areas elsewhere in the business; one has joined Whelan’s team and is being trained as a project manager.
The firm has thus been able to realise efficiency and insight gains without any upheaval.
Reassured has opted for a partnership-led approach in areas where it sees no competitive advantage in building proprietary tools, the AI quality solution being a prime example. “For something like an AI quality assurance tool, it doesn’t make sense to build it yourself,” Whelan argued. Instead, the business sought partner Voyc with “the skills and expertise and investment into that kind of technology”, co-innovating on new capabilities and configuring the platform to Reassured’s scoring model.
Alongside that, the company is working with a data and AI consultancy to accelerate capability while deliberately transferring knowledge back into the organisation. The partner, FOIL AI, has helped recruit a new analyst who is now on a two-year training programme where they will help develop his experience so he grows his skills as Reassured grows their capability.
“Everybody across the organisation has got an opportunity to apply for an apprenticeship,” Whelan said. Staff from finance, marketing and other business functions are training in data analytics, while colleagues in operations, HR and compliance are focusing on AI, learning how to spot opportunities and build business cases. Graduates are expected to cascade knowledge by running sessions for peers, reinforcing a data-literate culture rather than concentrating expertise solely in a central team.
Whelan describes Reassured’s appetite for change as unusually strong. But she is quick to stress that inclusion and communication are critical to sustaining that appetite.
The sponsors of major initiatives, she said, have “put themselves in the shoes of the people that are receiving that change” and worked to avoid the perception that transformation is simply something “being done to them”. In one recent AI project affecting the QA team, for example, she arranged for engineers to brief staff directly. Expecting only a handful of attendees, she was surprised when “the whole team turned up” and spent two hours engaging with the detail, buoyed by their positive experience with earlier AI deployments.
Reassured has also introduced “pre-mortems” at the start of major projects, convening a cross-section of stakeholders to identify potential risks and adoption barriers while there is still time to address them. That practice is, in part, a response to lessons learned from previous initiatives where enthusiasm led to rushed execution.
“We’re very excited. We want to prove the value and we want to get something done,” Whelan reflected. In hindsight, spending more time up front on roles, responsibilities and expectations would have “made it a little smoother”. The organisation is now putting greater emphasis on “a bit more upfront planning and communication in an environment setting”, while still accepting that some early-stage “friction” is inevitable in transformational work.
Perhaps the most ambitious piece of AI experimentation at Reassured is a non-disclosure detection tool in relation to incomplete or inaccurate medical disclosure. Aimed squarely at improving the experience for customers specifically in the event they need to claim on their policy.
Whelan frames the issue bluntly. Non-disclosure can arise where “a customer hasn’t disclosed a piece of medical information” that would later impact a claim, or where “an error is made completing the application.” “Neither of those are good customer outcomes,” she noted.
The new tool uses entity extraction to compare structured medical disclosures, the data on which a case was underwritten, with the unstructured transcript of the sales call. The aim is to “see whether there was a risk that we didn’t capture the disclosure correctly, or there was a partial disclosure that we didn’t explore enough.”
Reassured secured Microsoft funding for a proof-of-concept and has, Whelan said, “now got a tool that we believe we can identify those nondisclosure risks [with]. Where the customer said something not translated into the policy so we can go and resolve those and make sure that we put that right.”
Once the solution is in production, a second phase will look to mine patterns in the medical data itself, for example, where certain conditions commonly co-occur, and build prompts for agents. If two conditions are disclosed that typically correlate with a third, the system could nudge the agent to “just double check with the customer” whether that additional area needs to be explored.
The longer-term benefit is twofold: reducing the risk of declined claims and improving the quality and consistency of the initial underwriting conversation, particularly for less experienced agents who lack years of tacit pattern-recognition.
Reassured is also exploring sentiment analysis and markers such as hesitation within its QA tooling, to help distinguish between accidental omission and potential deliberate concealment by customers. That could support decisions to revisit cases and give customers “another chance just in case” to provide full information, particularly around sensitive topics.
Asked to look three to five years ahead, Whelan naturally focuses on AI, and specifically on generative and “agentic” AI as it applies to insurance workflows.
She cites current examples such as the use of generative models to summarise GP reports for underwriters as an important step forward, turning previously unstructured information into something more usable at speed. But she believes the real breakthrough will come when those capabilities are combined with orchestration layers able to handle “very complex workflows”, opening the door to far greater end-to-end automation in policy set-up, underwriting and claims.
Beyond that, she expects insurers and intermediaries will have to design, not only for human users, but for AI agents acting on their behalf. “We won’t just be building digital journeys for customers,” she suggested. “We’ll be designing it for AI agents that will be purchasing on behalf of customers as well.” Full automation of life cover purchase by third-party agents may be some distance away, but the trajectory is clear and the pace of AI adoption over recent years suggests that distance may shrink quickly.
For now, Reassured’s focus remains on building robust data foundations, experimenting with targeted AI applications such as QA analytics and non-disclosure detection, and investing heavily in people and culture so that the benefits of technology can be realised without sacrificing trust.
As Whelan puts it, if transformational change does not have its challenges, “then you’re probably not doing anything that’s that transformational.” The challenge for life insurance intermediaries, as Reassured’s experience illustrates, is to harness that productive chaos without losing sight of customers, colleagues or long-term value.
Away from the office, Whelan’s life is no less busy. With two school-age children, she jokes that a good part of her free time is spent on the road, ferrying them around to after school activities. When she is not operating as family chauffeur, she carves out time for yoga and reformer pilates, and heads outdoors whenever she can. A keen paddleboarder and dog owner, she often escapes to the Peak District, where walking in the hills provides a counterweight to the intensity of corporate transformation work and a space to reset her thinking.
