From Days to Hours: How Arch Insurance Automated Submission Intake

Speed is not a feature in commercial insurance. It is a competitive position. When an underwriter can prioritize and triage a submission first, quote first and bind first, conversion rates rise. When they are stuck waiting two or three days for a risk to be made decision-ready, the business goes elsewhere.

That is the reality Arch Insurance was navigating. One of the most respected specialty carriers in the market, Arch had the underwriting talent and appetite. What it lacked was an intake process fast enough to streamline operations and put that talent to work in time. Submissions were piling up while analysts manually performed data extraction from emails, attachments and supplemental documents – a time-consuming process that could stretch to three days before an underwriter ever saw a risk in decision-ready form.

The question Arch needed to answer was not whether AI-driven automation was possible. It was whether automation could be deployed at the speed, accuracy and configurability that a specialist carrier actually needs – across lines of business, geographies and both underwriting and claims processing intake – without dismantling workflows that were already functioning.

That question is what brought Arch to Cytora. And the results are worth being direct about.

The Problem That Actually Needed Solving

Manual submission intake is easy to describe and hard to fix. The surface-level problem – analysts reading emails, attachments, ACORD forms and spreadsheets to extract key data points and perform manual data entry – is obvious. The operational consequences are harder to see until they compound.

At Arch, the intake-to-decisioning cycle time was running two to three days. That stretched quote turnaround times to three to four days. Clearance performance sat at 44.9% within four hours – against an internal target of 80%. And because analyst capacity was consumed by repetitive manual effort in submission processing, submission volumes were effectively capped by headcount rather than appetite.

Underwriting teams were not underwriting. They were waiting. The bottleneck was not a skills gap – it was a structural one. Unstructured data arriving in inconsistent formats across document types was creating a processing backlog that no amount of analyst capacity could fully clear.

The challenge was finding a platform that could close it without creating a different set of problems – a long integration project, an expensive retraining cycle or a rigid schema that could not keep pace with how specialty lines actually evolve.

There is a version of automation that sounds good in a vendor presentation and fails in production. Arch needed to know they were getting the other version.

What Cytora Deployed at Arch

Arch selected Cytora™ to replace manual clearance with agentic AI – automatically extracting structured submission data from inbound insurance submissions and routing it directly into their downstream Submission Manager system, without disrupting existing underwriting workflows.

Four capabilities defined the deployment:

• AI-powered agentic digitization that requires no model training, enabling fast deployment across lines of business and geographies without waiting for labelled datasets – and eliminating the manual effort of traditional document processing
• Unified risk reasoning that structures data from submissions, internal systems and external sources into a single decision-ready view of risk, supporting faster risk assessment and higher data quality
• Configurable line-of-business schemas authored in natural language, giving Arch's team direct control to shape and expand how risk is presented – without requiring engineering support or manual data entry to accommodate new document types
• Agentic workflow automation designed, controlled and optimized by Arch's own team, with full auditability and human-in-the-loop validation checkpoints across underwriting and claims processing intake

The result is a submission intake process that moves from unstructured data to decision-ready risk in real time – with straight-through processing for high-confidence submissions and triage routing for those that need review. Each stage is auditable, configurable and controlled by the carrier.

The initial deployment covered four lines: Executive Assurance, Professional Liability, Healthcare and E&S Casualty. It went from contract signature to live production in under three months – a timeline that reflects how the platform is built: configurable rather than customized, and deployable without a long professional services dependency.

Arch's Strategic Analytics team was trained and certified as self-sufficient platform configurators. That detail matters. The measure of a platform deployment is not whether it works at launch – it is whether the customer can operate, adapt and expand it without ongoing external resources.

The Results

The metrics are concrete, so I will lead with them.

• Intake-to-decisioning processing times reduced by 70% – from two to three days down to two hours
• Quote turnaround moved to same-day, from three to four days
• 95% automated digitization accuracy achieved
• 87% classification accuracy, above target
• 90%+ clearance accuracy achieved in testing
• Arch's Strategic Analytics team is now fully self-sufficient – no external resources required to expand to new lines

The last point is as important as any of the accuracy metrics. Arch's internal team added cyber loss runs independently shortly after launch – demonstrating that the process automation model extends naturally as new lines or document types are added. That is not a proof of concept contained to a single deployment. It is an operational capability that compounds.

Faster processing times change more than just throughput. When decision-making moves from days to hours, underwriting teams have more time on actual risk assessment. That is where the competitive advantage materializes.

Allison Thornicroft, VP Business Solutions at Arch Insurance, put the strategic logic clearly:

"We know that if we are the first to quote we have a 50% higher conversion rate. Cytora has enabled us to reduce from days to a few hours the time it takes us to present decision-ready risks to underwriters."

– Allison Thornicroft, VP Business Solutions, Arch Insurance

That is not a statement about technology. It is a statement about competitive position. When submission intake is measured in hours rather than days, underwriters have more time on risk. When they have more time on risk, conversion improves. The operational change creates a commercial outcome.

What This Means for Commercial Carriers

Arch is a specialist carrier with the operational discipline to run a deployment like this well. But the underlying problem – submission volumes constrained by analyst capacity, clearance performance falling short of appetite, underwriting teams waiting for data that should already be structured – is not unique to Arch.

Most commercial carriers and MGAs have some version of this bottleneck in their underwriting process. Manual effort in intake is not just an operational inefficiency – it is a constraint on new business. Every submission that sits in a queue waiting for manual data extraction is a risk that a faster competitor might quote first.

The question is not whether to automate intake. The question is how to do it without creating a different kind of debt: a platform that requires constant model retraining, a schema that cannot evolve with the book, or an implementation that outlasts the business problem it was meant to solve.

The Arch deployment demonstrates what is possible when the platform is designed around configurability and self-sufficiency from the start. AI-powered agentic digitization without model training dependencies. Schemas authored in natural language by the business team. Underwriting workflows that the carrier controls, audits and optimizes. Production in under three months.

None of that is incidental. It is the architecture that makes the operational efficiency gains durable.

Moving Forward

Cytora is live across Arch's US and UK operations – in both underwriting and claims processing intake – across four lines of business, with an internal team capable of expanding independently.

The intake bottleneck that was constraining submission volumes at Arch is gone. What replaced it is a process that produces decision-ready risks in two hours, runs at 95% digitization accuracy, handles loss runs and supplemental documents without manual effort, and is operated entirely by the carrier's own team.

Not a pilot. Not a proof of concept. An operational capability that changes how underwriting teams spend their time and how the business competes for new business.

Read the Full Arch Insurance Customer Success Story →