Simon Glairy is a distinguished leader in the insurance technology landscape, currently serving as a Chief Technology and AI Officer with a career defined by high-stakes digital transformation. Having spent over a decade at Palantir, where he specialized in deploying software into some of the world’s most complex and crisis-driven environments, he brings a unique “forward-deployed” engineering philosophy to the financial sector. At the heart of his work is the belief that artificial intelligence should not be a superficial layer but a weight-bearing engine that fundamentally rearchitects how global organizations manage risk and serve clients. This conversation explores his transition from crisis management to insurance innovation, the mechanics of “hard mode” AI adoption, and the future of agentic commerce in global finance.
The following discussion synthesizes his perspectives on embedding technical teams within frontline operations, the evolution of digital client journeys through autonomous systems, and the imperative for organizations to move beyond experimental pilots toward integrated, end-to-end automated workflows.
You spent over a decade at Palantir building commercial business and scaling insurance partnerships. How does your experience with high-stakes, crisis-driven engineering translate to insurance workflows, and what specific steps are necessary to ensure new technology remains grounded in the needs of frontline brokers and operators?
My time at Palantir was defined by “open-heart surgery” on organizations during their most vulnerable moments, such as responding to the 2020 bushfires in Sydney or navigating the logistical chaos of COVID-19. These experiences taught me that impact is born from proximity; you cannot solve a problem from a boardroom when the frontline is in crisis. Translating this to insurance means adopting a forward-deployed engineering model where technical teams don’t just interview brokers but actually sit beside them to witness the “messiness” of the daily grind. To keep technology grounded, we must move past abstract strategy and engage in first-principles thinking, ensuring that every line of code addresses a specific friction point felt by the “doers” of the business. Ultimately, accountability must be measured by tangible results—like the speed of a placement—rather than the mere “activity” of launching a new tool.
Many organizations favor isolated AI pilots, yet there is a growing movement to embed AI directly into core, weight-bearing operations. What are the primary trade-offs of this “hard mode” approach, and how do you manage the technical risks when rearchitecting a global organization’s fundamental engine?
The “easy mode” of isolated AI labs and minimum viable products is comfortable because it carries low risk, but it rarely changes the fundamental trajectory of a company. Choosing “hard mode” means embedding AI into productive workflows that must bear real weight across the value chain, which naturally increases the stakes of any technical failure. We manage these risks by ensuring there is zero separation between the people designing the strategy and those executing the code. By building cross-functional platforms that unite colleagues with AI in core propositions—such as premium placement and claims—we create a feedback loop that catches errors in real-time. This approach requires a cultural shift where the organization accepts that the current way of operating is never “enough” and views every failure as a data point for refinement.
Speed is a critical differentiator in financial services, particularly for complex aviation quotes that typically take days. Could you walk through the step-by-step process of using AI to connect client interactions with carrier data, and what metrics best demonstrate the impact of reducing these turnaround times?
We developed Project LeftSeat specifically to tackle the agonizingly slow pace of aviation quoting by creating an end-to-end engine that bridges the gap between client data and carrier requirements. The process begins with AI agents capturing and interpreting initial client interactions, which are then enriched with real-time data and routed through specific carrier integration rules automatically. By removing the manual friction of data entry and verification, we have seen turnaround times move from days to mere minutes. The best metrics for success here aren’t just time-savings, but the “alpha” generated through higher volume and the improved consistency of outcomes for our clients. It transforms the professional’s role from a data processor to someone wearing a “cybernetic suit,” where their human judgment is amplified by the speed of the machine.
Digital client experiences are evolving from simple portals to autonomous systems capable of “agentic commerce.” How do AI agents interpret communications to proactively coordinate services, and what practical strategies can firms use to ensure these automated workflows amplify rather than replace human judgment?
The shift toward agentic commerce means moving beyond a “slick portal” to a system that understands the intent behind a communication and proactively moves the work forward. These AI agents interpret unstructured data from emails or calls, enrich it with relevant context, and then autonomously coordinate across various systems to handle renewals or advisory tasks. To ensure this amplifies rather than replaces humans, we design these systems as an “Iron Man suit” for our brokers, where the AI handles the heavy lifting of coordination while the broker retains the final decision-making authority. Practically, this involves building the AI surface as a partner that presents the best options for due diligence and selection, allowing the professional to focus on high-level relationship management.
Innovation often fails when there is a significant gap between strategy and execution. How can leadership empower technical teams to work directly in the field alongside business leaders, and what methods do you use to foster an organizational culture where truth-seeking is rewarded over mere activity?
Leadership must grant technical teams the explicit permission and encouragement to challenge the status quo and “spend time at the coalface” where value is created. This means dismantling the silos between the engineering office and the brokerage floor so that teams share joint ownership of both the development and the eventual business outcome. We foster truth-seeking by rewarding results rather than the number of hours worked or the number of features launched. When innovation becomes part of the DNA, failure is no longer a stigma but an essential part of the engineering process that leads to more robust systems. It requires a relentless ambition to re-earn our “fintech medal” every single day by proving that our tech actually makes the business run better.
Bringing AI into core operations requires treating it as a bottom-up engineering exercise rather than an add-on. For organizations looking to move past the tinkering phase, what is the best way to prioritize systems that coordinate end-to-end workflows, and how should they design for compliance and accountability?
Organizations need to stop theorizing and start by identifying where human judgment meets repetitive inefficiency in their current daily processes. The priority should always be on systems that can orchestrate a full workflow—from initial contact to final policy issuance—rather than just automating a single task like document scanning. Designing for reality means that compliance, permissions, and accountability are baked into the architecture from day one, not treated as an afterthought or an “add-on.” My advice is to find partners who are actually doing the work, link arms, and build momentum through small, weight-bearing successes that prove the system’s reliability. It’s a rigorous, bottoms-up exercise that demands a willingness to take on the hardest problems first.
What is your forecast for AI in the financial sector?
I believe we are approaching a “Google Search 1990s” moment where AI surfaces will become the primary way clients interact with financial services. In the near future, the most successful firms will be those that have transitioned from responsive systems to proactive ones that orchestrate due diligence, selection, and execution autonomously. We will see the rise of “agentic orchestration” moving first within enterprises and then across them, fundamentally changing how risk is placed and managed globally. The organizations that decide to operate on “hard mode” today, building deep AI foundations into their core engines, will be the only ones capable of meeting the inevitable spike in client expectations for near-instant, high-quality advisory. Those who continue to treat AI as a peripheral pilot project will simply be left behind as the pace of digital commerce accelerates beyond their manual reach.
