Certainty in an uncertain world

An Interview with Rob Lewis (CEO INTX) by Sean Murphy

Somewhere along the chain that runs from a broker’s first quote through to the reinsurer’s recovery, a typical US fronting carrier with two billion dollars in annual premium loses between five and ten percent of what it should be collecting back, by industry estimates.

Anywhere from 100 to 200 million dollars a year, draining out through under-administered treaties, mis-allocated layers, sliding-scale commissions that nobody has reopened, and spreadsheets that do not quite reconcile. The money does not disappear dramatically. It bleeds.

For Rob Lewis, the chief executive of Austin-based INTX Insurance Software, that bleed is the most visible symptom of a quieter, more structural problem. What every carrier and risk manager actually wants, he says, is contract certainty: knowing exactly what is covered, by whom, for how much, at every point in the lifecycle, in real time. Sounds easy, but after two decades of insurance software investment, he feels that industry still has not delivered on this.

The lifecycle in question is long, and full of handoffs. A risk enters through a broker portal, moves into underwriting where coverages are priced and terms set, becomes a bound policy, generates premium that has to be billed and accounted for, and may produce a claim that has to be triaged, reserved and paid. In parallel, the same risk has to be allocated across reinsurance treaties or facultative contracts, with ceded premium, recoveries, and dynamically adjusting commissions tracked over the life of the policy. Each handoff is a place where certainty leaks away. Every system boundary, every spreadsheet, every manual reconciliation is somewhere that the actual contract on the page has the potential to be transformed into something slightly different and in this industry small changes can have massive repercussions.

In a typical carrier, those handoffs sit across different systems and different teams: an underwriting platform that does not speak to claims, a billing tool bolted in from somewhere else, a reinsurance department reconciling its position to the policy book in Excel at the end of each month. Information moves between them slowly, by export and re-import, by email attachment, by a junior analyst rebuilding a workbook. By the time a fast-moving event arrives, whether wildfire, flood or regulatory enquiry, the picture of who owes what to whom is rarely current.

Lewis and his brother bought an insurance carrier in Mozambique in 2001, could not find software that worked, and ended up commissioning their own. That platform became Kindle Insurance Technologies, sold into emerging markets and the UK for two decades, and now lives on as the core of INTX. The company runs three products under a single architecture: a core insurance operating system spanning the full policy lifecycle from broker portal through claims and regulatory filing; an assumed-reinsurance system called INTX Re; and ReinsureConnect, the ceded reinsurance product launched at ITC Vegas last October. Lewis frames the offering not as another piece of modernised core software but as a complete operating system for, in his words, “managing the economics of an insurance company.”

Reinsurance is where the certainty problem becomes most acute, and where Lewis’s commercial story has been most interesting. INTX has nine MGAs, reinsurers and carriers on the core platform, but its reinsurance conversations include five carriers writing more than a billion dollars in annual premium, with tier-one names among them. The pull is software that takes the complexity seriously: ingesting policy data, allocating layers across proportional and non-proportional structures, handling the sliding-scale, loss-corridor and override commissions that change shape as a treaty’s loss ratio moves, recalculating recoveries dynamically as claims incur, and refusing to bind a risk the reinsurance programme will not support. 

The reason the gap has stayed open so long, he argues, is that Excel is as good and sometimes better than many of the current systems on the market. Industry research the company commissioned found that 72 percent of insurers still rely on Excel or internally built tools to manage critical workflows, with more than half of policy workflows requiring manual intervention. The spreadsheet is powerful, cheap, and infinitely flexible. It is also the enemy of certainty: unauditable, easily altered, and impossible to extract cleanly when something goes wrong..

New technologies are not always better and Lewis is careful not to overplay the benefits of AI. INTX is not trying to compete with Claude or ChatGPT; its job is to deliver clean, real-time, auditable data underneath models carriers are mostly building themselves, because AI built on uncertain data simply industrialises the uncertainty. “If you let that animal loose,” Lewis says of an unsupervised AI model, “you would be amazed at how quickly it can start accumulating risk.”

The insurance industry is built on managing risk in uncertain scenarios, so when it can find certainty anywhere, it does its best to protect it. Alignment between what a policy says, what a system records, what a reinsurance treaty will pay, and what arrives in the bank is precisely the kind of certainty worth protecting. 

The case Lewis is making, across both his core platform and his reinsurance products, is that the industry can finally have it.