Rachel Oliver (pictured), COO at Moonrock Insurance, did not set out with insurance as an obvious destination. More than two decades on, after a career spent moving across operations, claims, planning and underwriting-adjacent roles, she now sits at the centre of Moonrock’s effort to sharpen underwriting discipline, improve broker service and prepare a specialist aviation MGA for a more AI-native future.
Oliver’s route into the market was less a deliberate calling than a well-timed opportunity. She recalled that, when she entered the profession, insurance was not among the most coveted graduate destinations. “Over two decades ago, insurance wasn’t really the forefront” of the graduate market, she said, noting that finance and larger brand-name schemes attracted more attention at the time.
What changed her trajectory was exposure. Joining one of the few insurance graduate schemes then available, she rotated across the business and developed a broad operational understanding rather than a narrow specialism. She said she was “exposed as a graduate” to insurance, operations, business planning, claims and wordings, and that this cross-functional grounding remains central to how she works today.
That breadth has shaped both her self-description and her usefulness inside transformation programmes. Oliver said she thinks of herself as an “internal plumber”, someone who understands how the moving parts connect and can keep processes flowing across the organisation.
At Moonrock, the specialist focus is on unmanned aircraft systems (UAS), and other emerging aviation risks, and Oliver’s definition of digital transformation is conspicuously unsentimental. It is not, she argued, a matter of buying “shiny new tools”, but of removing friction from the insurance value chain for brokers, customers and underwriters alike.
That matters especially in a segment where clients expect speed, clarity and precision. Drone operators, she observed, do not want administrative drag; they “simply want to fly safely and as soon as possible, as insurance is a mandatory requirement for commercial operators”. Brokers, meanwhile, want quotes and policy documentation without delay, and underwriters need better information arranged in ways they can use.
Moonrock’s digital story was already well underway long before Oliver arrived. Although she joined the business less than two years ago, the foundations had been laid almost a decade earlier, when Moonrock launched a click-and-bind broker portal for drone insurance. At the time, it was a genuinely pioneering move: the first fully online click-and-bind offering in a highly specialised niche. What has shifted more recently, she says, is not the ambition, but the speed and pace of evolution.
Oliver points to two programmes in particular. The first is the quote-to-bind journey, designed to support both broker and customer through a cleaner digital process and ensure data enters the business in a more usable form. The second is an underwriting workbench that brings together pricing, workflow, documentation and exposure data on the specialty book.
The significance of those initiatives, in her account, is not merely that they accelerate transactions. Their greater value lies in creating structured datasets robust enough to support AI use cases properly. Without that preparation, she suggested, any AI overlay would be cosmetic. “Without that groundwork,” she said, “the AI would just be superficial.”
On technology partners, Oliver’s criteria are practical rather than evangelical. After years of seeing vendor presentations, the issue, she said, is rarely whether a system looks impressive in isolation. The real test is whether it can be translated into something meaningful for the business, and whether the provider understands the operating realities of the London market.
That preference has led Moonrock towards providers with direct insurance operational experience, including people who have worked inside the industry before moving into technology. In Oliver’s view, that shortens the journey from requirement to implementation by reducing misunderstanding between buyer and supplier.
She later identified MEA as the firm’s AI ingestion partner, alongside some internally built capabilities.
Pressed on the business case, Oliver’s answer was concise: “speed, loss ratio control, broker experience, operational resilience”. Those are the outcomes she is trying to improve, but the enabling condition is visibility.
She reached for an aviation analogy to make the point. Just as a drone operator needs the right information about direction, altitude and movement in order to act safely, so management needs the right data in front of it in order to make timely strategic decisions. The ambition is to make the organisation more agile by using information well, not simply to digitise inherited processes.
That is a familiar claim in transformation circles, but Oliver gives it some force. Digital transformation, she said, is “not necessarily a technology problem. It’s a human one.” In a fast-moving UAS market, no organisation can rewire workflow, data foundations, broker experience and tooling all at once without overwhelming its people.
Her focus therefore has been on sequencing change, pacing adoption and keeping teams aligned on the reason for doing it. Resistance, where it appears, is less about fear of redundancy than about the ordinary discomfort of having to work differently. The objective is to augment people rather than replace them.
If there is one area where transformation optimism collides with reality, it is data. Oliver described “data readiness” as the hardest part of the exercise, requiring clean, standardised and governed information before technology can fulfil its promise. The effort involved, she implied, tends to exceed initial expectations even when one knows in theory that it will be painful.
Moonrock benefits from some structural advantages. Unlike many incumbents, it has not been trapped by deeply embedded legacy systems, allowing it to build much of its data framework on a greener field. In addition, as an early entrant to the market, the business has also accumulated a depth of highly industry-specific data that few others can match.
Against that backdrop, Oliver’s account of AI is notably grounded. At present, she sees it as valuable in submission ingestion, triage and pricing checks, especially because every submission arrives in different form. AI helps convert that variability into structured information and frees underwriters to spend more time on judgment-led decisions rather than repetitive administrative work.
That distinction between automation and judgment matters. Oliver does not frame AI as a substitute for underwriting expertise, but as a way to redeploy it more intelligently.
What makes this corner of the market unusual, in her telling, is the pace at which underlying risk changes. In drone and emerging aviation lines, “the risks are evolving monthly, not annually”, which means the transformation agenda must move faster than in more established books.
That has implications for the type of data insurers need to ingest and interpret. Operational flight data, patterns, payload characteristics and environmental conditions all matter more in a segment tied so closely to technological advancement.
Oliver also expects the next few years to bring more AI-native underwriting, richer use of real-time sensor and geospatial data, and more dynamic pricing. In a data-rich environment such as UAS and aviation, she said, insurers will increasingly be able to use information on mission profiles, aircraft class, geospatial environment and operator behaviour to refine underwriting decisions.
She is careful, however, to distinguish between AI-assisted and AI-orchestrated models. The industry’s direction of travel, in her view, is towards “automated intelligence” rather than a thin layer of assistance sitting atop traditional processes.
Some of the most interesting applications sit where underwriting and claims data meet. Oliver noted that drones effectively carry the equivalent of a “black box”, allowing insurers to reconstruct what happened after an incident and learn from that information.
That same data can improve underwriting. She highlighted the importance of autonomous operations, arguing that greater automation in drone use could reduce future losses. Geofencing and other operational controls help, but human fatigue still introduces error; more fully automated systems and operations may eventually lower claims incidence.
For readers less close to the sector, Oliver also clarified two terms now likely to surface more often in specialist aviation coverage: BVLOS means “Beyond Visual Line of Sight”, while eVTOL refers to “Electrical Vertical Take-Off and Landing”, the latter encompassing future passenger and cargo systems.
If Oliver sounds deeply immersed in operational complexity, her idea of unwinding is almost aggressively simple, “I don’t require rituals or gadgets, I literally just close my eyes and just stop thinking,” she said.
She also described life at home with teenage children, suggesting that, after a day spent steering change, systems, people and aviation risk, silence may be its own form of luxury.
