Cytora: IDP Software Vendor
On This Page
AI-powered digital risk processing platform for commercial insurance offering submission digitization and underwriting automation through generative AI workflows.
Overview
Founded in 2012 as a University of Cambridge spinout, Cytora develops AI for commercial insurance underwriting. The London-based company raised $41.43M before being acquired by Applied Systems in September 2025 for an undisclosed sum. Richard Hartley, CEO and co-founder, has consistently framed the company's mission around a persistent market inefficiency: carriers fail to return quotes 30-50% of the time because underwriters spend 40-50% of their time reading and interpreting submission data rather than making risk decisions.
The Applied Systems acquisition immediately accelerated product integration. Within weeks, Cytora launched Epic AutoFill and Epic Submissions Manager at Applied Net 2025, embedding its AI directly into Applied's Insurance AI suite. The company has since pursued a data ecosystem strategy, accumulating proprietary enrichment integrations as a switching-cost barrier rather than competing on AI features alone.
In March 2026, Cytora crossed a significant product threshold with the launch of Cytora Autopilot, an agentic AI capability that executes underwriting and claims workflows autonomously without manual intervention. Hartley described the shift directly: "The industry has reached the limits of static digitization. We've seen enormous progress turning submissions into structured data, but workflows themselves have remained largely manual. Autopilot changes this dynamic by enabling workflows that understand context dynamically, respond to new information and execute autonomously as the full picture of a risk evolves."
Alongside Autopilot, Cytora has added three data enrichment partnerships since January 2026: Climatig for climate hazard analysis, The Warren Group for U.S. property intelligence drawn from a database built continuously since 1872, and Ideal Postcodes for rooftop-level address validation and geocoding. In March 2026, the Business Intelligence Group awarded Cytora the 2026 Artificial Intelligence Excellence Award in the insurance product category, citing customer outcomes including premium uplift and improved risk selection.
The platform uses large language models pretrained for commercial insurance with zero training required, supports 140+ languages, and holds ISO 27001 and ISO 42001 certification. Enterprise customers include Chubb for global claims automation and Markel, which reported a 113% productivity increase from implementation. Both figures originate from Cytora's own materials; no independent analyst assessments from Gartner, Forrester, or Celent are available in current sources.
How Cytora processes documents
Cytora's architecture targets the submission pipeline at three sequential stages: ingestion, enrichment, and routing. The March 2026 Autopilot launch adds a fourth layer: autonomous workflow execution that operates continuously as new information arrives.
At ingestion, the Document-to-Data layer accepts ACORD forms, PDFs, Excel files, CSV files, emails, and images, converting them into structured data using insurance-pretrained LLMs that require no training datasets. Fuzzy matching via the Ideal Postcodes integration automatically corrects common address errors at this stage, enabling property-level risk assessment before the submission reaches an underwriter.
At enrichment, the Verify API augments extracted submission data with external sources. The Warren Group's national real estate database injects ownership structures, transaction histories, mortgage encumbrances, and pre-foreclosure signals for commercial property lines. Climatig's hazard models add river flood, severe wind, and wildfire risk scores for climate-exposed assets. Ideal Postcodes contributes Rooftop Geocodes, UK and international address data, and Unique Property Reference Numbers (UPRN) for building-level pricing precision. Juan de Castro, COO, described the intent: "Precise property and ownership data is critical for modern commercial underwriting. This enables our clients to automate the validation of property assets and ownership structures instantly, allowing underwriters to focus on technical risk evaluation rather than manual data gathering."
At routing, the Risk Flow Engine applies configurable business rules to direct decision-ready packages to underwriters, automating triage without manual intervention. Epic AutoFill pre-fills carrier document fields automatically; Epic Submissions Manager provides a unified workspace for tracking submission-related emails and attachments within Applied Epic.
Autopilot operates across all three stages simultaneously. The system aggregates data from emails, documents, calls, and submissions, then executes workflows autonomously as new information arrives. It maintains persistent workflow context across communications, linking data that arrives at different times, and provides cross-portfolio risk visibility. Critically for regulated environments, Autopilot includes explainable agentic reasoning with auditable records of every workflow decision. Cytora claims turnaround times fall from hours or days to minutes with Autopilot active. No independently verified benchmarks support this figure.
The Warren Group integration is structurally significant because it operates before the underwriter sees the submission, not as a post-decision reference tool. David Lovins, CEO of The Warren Group, noted: "By delivering our national real estate and transaction intelligence as AI-ready data within the Cytora platform, we are equipping commercial insurers with structured, scalable insights designed for advanced analytics and automation."
No source names the specific LLM models underlying the platform, provides release dates for the "advanced LLM capabilities" referenced in Cytora's press materials, or offers benchmark data. The competitive significance of Cytora's AI layer relative to peers such as Indico Data, SortSpoke, or Convr cannot be assessed without it.
Use cases
Commercial insurance submission processing
Cytora automates submission intake for commercial P&C risks, addressing the market inefficiency where carriers fail to return quotes 30-50% of the time due to manual processing costs. The platform digitizes incoming submissions, enriches them with external data, and applies business rules for automated routing to reduce the volume of submissions that stall before reaching an underwriter. With Autopilot, the system now executes follow-up actions autonomously when submissions arrive incomplete, rather than queuing them for manual review.
Commercial property underwriting
The Warren Group partnership extends the enrichment layer specifically for commercial property lines. Ownership structures, transaction histories, mortgage encumbrances, and pre-foreclosure indicators drawn from a database built continuously since 1872 are injected automatically at the submission stage, replacing manual property research. The Ideal Postcodes integration adds rooftop-level geocoding and UPRN data, enabling building-level pricing accuracy. No quantified metrics for premium leakage reduction or time-to-quote improvement have been disclosed across any source.
Claims automation
Through the Chubb engagement, Cytora enables automatic digitization of claims documents, eliminating manual intervention and creating scalable AI-native claims flows across multiple markets and lines of business. Autopilot extends this by maintaining persistent context across the full claims lifecycle, linking documents, emails, and calls that arrive at different times into a single coherent workflow.
Climate risk compliance
The Climatig partnership provides automatic geolocation of assets and climate risk assessment covering river flood, severe wind, and wildfire hazards to support TCFD regulatory requirements for commercial insurers. This positions Cytora beyond document processing into the risk analytics layer that regulators increasingly require.
Carrier-agency workflow automation
Autopilot's support for agentic collaboration across carrier-agency relationships addresses fragmented data flows between brokers and carriers. By automating data exchange and workflow completion across organizational boundaries, the system targets a coordination bottleneck that has historically required manual follow-up on both sides.
Technical specifications
| Feature | Specification |
|---|---|
| Core platform | Risk Flow Engine, Document-to-Data Platform, Cytora Autopilot |
| AI technology | Pretrained LLMs for commercial insurance, zero training required |
| Language support | 140+ languages |
| Document types | ACORD forms, PDFs, Excel, CSV, emails, images |
| Enrichment integrations | The Warren Group (U.S. property intelligence), Climatig (climate hazard data), Ideal Postcodes (address validation, UPRN, rooftop geocoding), Verify API |
| Agentic capabilities | Persistent workflow context, autonomous execution, cross-portfolio risk visibility, auditable reasoning |
| Integration | Applied Epic (Epic AutoFill, Epic Submissions Manager), API-based data exchange |
| Certifications | ISO 27001, ISO 42001 |
| Deployment | Cloud (Applied Systems infrastructure) |
Company information
Headquarters: London, United Kingdom
Founded: 2012 (Cambridge spinout)
Funding: $41.43M total (pre-acquisition)
Acquired: September 2025 by Applied Systems
Resources
Related capabilities: Generative AI · Data Extraction · Integration and Workflow · Agentic Document Processing
Related guides: Insurance Claims Processing · Real Estate Document Processing · Document Workflow Automation
Compare vendors: Indico Data · SortSpoke · Convr · Mea Platform