Insurance runs on a decision that has barely changed in a hundred years. An underwriter receives a submission, digs through a stack of documents, weighs the risk against the carrier's appetite, and decides whether to quote, how to price it, and on what terms. That decision is where an insurer makes or loses money. It is also, for most carriers, the single biggest bottleneck they own.
A New York company called Sixfold has spent the last three years building artificial intelligence aimed squarely at that bottleneck. In June 2026 it put a name to the work: the AI Underwriter. This is a deep look at what Sixfold has built, why some of the largest carriers in the world already run it, and, just as important, what it actually takes to get value from a tool like this once the contract is signed. Because the second part is where most enterprise AI quietly succeeds or fails.
The problem Sixfold set out to solve
Underwriting has a quiet productivity crisis. The work is expert work, but a large share of an underwriter's day is not spent on judgment. It is spent on retrieval and reconciliation: opening loss runs, cross-referencing statements of value, reading medical records and prescription histories, checking each submission against pages of guidelines, and re-keying the results into a policy administration system. The expensive expertise sits idle while the expert does clerical work.
Sixfold's founders had seen this from the inside. CEO Alex Schmelkin spent more than two decades in insurance and previously co-founded Unqork and Cake & Arrow. He started Sixfold in 2023 with COO Jane Tran and CTO Brian Moseley, a former Head of Developer Experience at American Express. Tran, a former JP Morgan and Marsh executive, was also a founding team member and COO at Unqork. Their read on the market, described on the company's About page, was that underwriters were stuck with poor technology while pouring scarce expertise into tasks a machine could do. They believed both problems were fixable, and that the fix was the same tool.
What the AI Underwriter actually does
Sixfold is not a chatbot bolted onto a carrier's inbox. It is a workflow agent that operates on real submissions and produces underwriting-grade output. The company splits its product into two lines that map to how insurers are actually organized.
On the Property & Casualty side, the system reads complex commercial documents such as loss runs and statements of value, scores the risk with explainable and source-cited signals, evaluates how the account fits the carrier's appetite and portfolio, and recommends the next best action. Every signal it surfaces is traceable back to the document it came from, which matters enormously in a regulated business where a decision may have to be defended years later.
On the Life & Health side, the same core capability is pointed at a different set of documents. It summarizes medical records and prescription histories, surfaces impairments and mortality factors with the supporting evidence attached, and enables faster and more consistent case decisions with a full audit trail.
The common thread is explainability. Sixfold does not ask an underwriter to trust a black box. It shows its work, cites its sources, and leaves the final call with the human. That design choice is the reason a compliance-heavy industry has been willing to let it near the risk decision at all.
In June 2026 Sixfold packaged this into what it calls the AI Underwriter, an agent that can carry a submission through evaluation and, in some configurations, all the way to straight-through quote and bind. Trade coverage of the launch, including The Insurer and Reinsurance News, framed it as one of the first credible attempts to automate the full underwriting motion rather than just a slice of it.
Who benefits, and the proof in the carrier logos
The direct beneficiaries are clear. Underwriters get their time back for judgment instead of paperwork. Carriers get faster, more consistent decisions and more capacity from the same team. Brokers and their clients get quicker quotes. And the business gets more gross written premium per underwriter without adding headcount.
Insurance is a conservative buyer, so the customer list is the most persuasive evidence. Sixfold's clients include Guardian, Zurich North America, AXIS, Skyward Specialty, Generali Global Corporate and Commercial, and New York Life. Together, the carriers running Sixfold represent roughly $270 billion in gross written premium.
The reported results are the kind that get a project renewed. According to figures the company publishes in its reports, customers have seen processing times fall by between 50 and 97 percent, quote-to-bind ratios rise by 15 percent or more, and gross written premium per underwriter climb by as much as 30 percent. Guardian's Head of Individual Markets has publicly credited the tool with cutting underwriter review time in half. Perhaps the most telling number is adoption: Sixfold reports a 90 percent-plus average adoption rate among its clients, which is rare for enterprise software and rarer still for AI tooling that asks experts to change how they work.
The company also says it has processed more than 1.5 million submissions across 50-plus lines of business on four continents, and it was featured at OpenAI DevDay 2025 for crossing 100 billion tokens processed. Skyward Specialty went a step further than most customers and entered a formal partnership to advance AI-powered underwriting together.
The money and the backing
In January 2026 Sixfold raised a $30 million Series B led by Brewer Lane, bringing its total funding to roughly $51.5 million across three rounds. The round was covered by outlets including FinTech Global and FF News.
The more interesting detail than the dollar figure is who else showed up. The round carried strategic backing from Guidewire, the dominant core-systems platform in property and casualty insurance. When the company whose software already sits at the center of most carriers' operations invests in an AI underwriting agent, it is a signal about where the industry's plumbing is heading. It also hints at how Sixfold expects to reach carriers: not as a rip-and-replace, but as an intelligent layer that plugs into the systems insurers already run.
Recognition has followed the funding. Sixfold was named Best Underwriting Solution at the 2025 British Insurance Technology Awards, made the CB Insights Fintech 100, went through Lloyd's Lab Cohort 12, and won the Zurich Innovation Championship. Its headquarters sit at 121 E 24th Street in New York, with a second office inside Lloyd's of London.
Why this matters beyond insurance
Sixfold is a clean example of a pattern showing up across enterprise AI. The winners are not the general-purpose tools that promise to do everything. They are the vertical, deeply specialized systems that go narrow, learn the rules of one high-stakes decision, and earn trust by showing their work. Underwriting is a near-perfect target for this: high volume, expensive expertise, document-heavy inputs, clear economic value in every point of improvement, and a hard requirement for explainability.
It also illustrates the real bottleneck in enterprise AI, which is not the model. It is adoption. The reason Sixfold's 90 percent adoption number stands out is that most enterprise AI never gets there. This is the same shift reshaping how work itself gets delivered, the move away from renting labor by the hour and toward buying outcomes that are measured and proven. The intelligence is increasingly a commodity. The scarce, valuable thing is the ability to turn it into a result inside a real organization. That is a theme we have written about as the future of work, and AI underwriting is one of its clearest live examples.
Buying the tool is easy. Adoption is the project
There is a common assumption that modern AI is plug-and-play. Point it at your data, and it works. For a horizontal chatbot, maybe. For an AI underwriter that has to make or inform a regulated risk decision, no. Getting real value is a project, and it has at least five distinct pieces of work, none of which the software does for you.
1. Appetite and guideline configuration. An AI underwriter is only useful if it reflects your appetite, not a generic one. Every carrier has its own risk appetite, guidelines, thresholds, and exceptions, often differing by line of business and region. Encoding that, testing it against real historical submissions, and tuning it until the recommendations match what your best underwriters would do is careful, expert work. It is also what determines whether the output gets trusted or quietly ignored.
2. Document ingestion and data plumbing. Underwriting inputs are messy. Loss runs, statements of value, ACORD forms, broker emails, medical records, and prescription histories arrive in dozens of formats. Getting the system to read them reliably means handling every document type your submissions actually contain, connecting to the systems and inboxes where they land, and validating the extraction against ground truth. This is unglamorous integration work, and it is where timelines slip when it is underestimated.
3. Integration into core systems and workflow. An AI recommendation that lives in a separate tab does not get used. Value depends on the output landing inside the systems underwriters already work in, whether that is a policy administration system, a core platform like Guidewire, a rating engine, or a workflow queue. Straight-through processing only works when that integration is genuinely done, not demoed. This is often the single largest slice of the effort.
4. Underwriter change management and trust. This is the piece that gets skipped, and the one that decides success. Underwriters are experienced professionals being asked to trust a machine on the core of their craft. If they do not trust it, they will work around it, and adoption dies no matter how good the model is. Building that trust takes explainability they can see, a rollout that starts where the wins are obvious, training that respects their expertise, and a feedback loop so their corrections visibly improve the system. The carriers reporting 90 percent-plus adoption did not get there by accident. They treated change management as a real workstream.
5. Measurement and continuous tuning. The value case has to be proven in the carrier's own numbers, not the vendor's marketing. That means baselining cycle time, hit ratio, and premium per underwriter before go-live, then measuring the same metrics after, on a defined segment, and tuning based on what the data shows. Without this loop, you cannot prove ROI to the executives who funded the project, and you cannot defend the next phase of the rollout.
Why carriers stall, and it is not the software
Notice what none of the five require: a better AI model. The bottleneck is almost never capability. It is capacity. Getting an AI underwriter fully live is a multi-month, cross-functional effort spanning underwriting, IT, data, and compliance, and most carriers do not have a spare team that understands both insurance workflows and AI deployment sitting idle waiting to run it. The vendor's customer success team is stretched across every account. The carrier's own people are busy underwriting. So the project stretches, the pilot lingers, and the value the business case promised arrives late or not at all.
This is not a Sixfold problem or an insurance problem. It is the defining problem of enterprise AI right now, and it is exactly why the old models for adding capacity fall short. Hiring a permanent team for a one-time deployment push is slow and expensive. Traditional outsourcing or offshoring rents you bodies by the month and leaves you managing them. Neither is built to deliver a specific outcome on a deadline.
Closing the gap faster
The carriers that win with AI underwriting treat implementation as the actual project and resource it properly. That usually means bringing in a delivery layer that sits behind the vendor's product and the carrier's team and owns the last mile: configuring the appetite and guidelines, building the document pipelines, doing the core-system integration, running the underwriter change management, and standing up the measurement loop that proves ROI.
This is exactly the work a Virtual Delivery Center is built to deliver. AiDOOS runs elastic delivery pods that plug in alongside a product like Sixfold and the carrier's own staff, take on the deployment and adoption work, and get the system live and proving value in weeks rather than quarters. The vendor keeps its focus on the product. The carrier gets to the outcome faster without building a permanent team for a one-time push. And because the work is scoped as delivered outcomes rather than bodies rented by the month, the carrier pays for progress, not effort.
The sensible way to start is small and provable: pick one line of business or one segment, get the AI underwriter fully live on it, measure the before and after on real numbers, and use that proof to fund the broader rollout. Prove the loop on a slice, then scale it with confidence.
The bottom line
Sixfold has taken one of insurance's oldest and most valuable decisions and rebuilt the workflow around it with AI that explains itself. The carrier logos, the retention-grade results, and the strategic money from Guidewire suggest it is early but real. The technology, from vendors like Sixfold, is ready.
The question every carrier should ask before signing is not whether the AI works. It is who is going to do the work to make it work here, and how fast. Answer that well, and the business case comes true. Answer it by assuming adoption will happen on its own, and the tool joins the long list of enterprise AI that was capable but never quite delivered. The model is no longer the hard part. Getting it adopted is, and that is a solvable problem.
If you are deploying an AI underwriter or any complex enterprise AI, AiDOOS provides the delivery pods that get it live and proving ROI. To explore the product itself, visit Sixfold. For more on where enterprise delivery is heading, read our North Atlantic Briefing.